SPB Git
spboucher.ai

spb/wp44_uqo

Public

WP44: category-level actor and impact indicators

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Simon-Pierre Boucher committed (Nov 4, 2026) parent cc7bdc1 Browse files

64 changed files +2,163 −547

64 changed files
Unified Split
modified bibliometrics/04_diversity_cagr.py +1 −1
@@ -12,7 +12,7 @@ import argparse, os, json, numpy as np, pandas as pd
12 def main(): 12 def main():
13 ap = argparse.ArgumentParser(); ap.add_argument("--root", required=True); ap.add_argument("--B", type=int, default=500); a = ap.parse_args(); R = a.root 13 ap = argparse.ArgumentParser(); ap.add_argument("--root", required=True); ap.add_argument("--B", type=int, default=500); a = ap.parse_args(); R = a.root
14 out = f"{R}/results/bibliometrics"; cc = pd.read_csv(f"{R}/results/graph/category_counts_yearly.csv"); cc = cc[cc.year <= 2025] 14 out = f"{R}/results/bibliometrics"; cc = pd.read_csv(f"{R}/results/graph/category_counts_yearly.csv"); cc = cc[cc.year <= 2025]
15− piv = cc.pivot_table(index="year", columns="category_id", values="n_primary", aggfunc="sum", fill_value=0); cats = list(piv.columns) # aggfunc="sum": the file has one row per (year, period); the default mean halved the 2018-2022 levels (fixed 2026-11-04, Part IX writer) 15+ piv = cc.pivot_table(index="year", columns="category_id", values="n_primary", aggfunc="sum", fill_value=0); cats = list(piv.columns) # aggfunc="sum": the file has one row per (year, period); the default mean halved the 2018-2022 levels (fixed 2026-11-03, Part IX writer)
16 rows = [] 16 rows = []
17 for c in cats: 17 for c in cats:
18 b0, b1 = piv.loc[2018, c], piv.loc[2022, c]; p0, p1 = piv.loc[2023, c], piv.loc[2025, c] 18 b0, b1 = piv.loc[2018, c], piv.loc[2022, c]; p0, p1 = piv.loc[2023, c], piv.loc[2025, c]
added bibliometrics/05_category_impact.py +28 −0
@@ -0,0 +1,28 @@
1+#!/usr/bin/env python3
2+"""WP44 — category-level actor and impact indicators (after classification): per primary category and year —
3+works, share with a China affiliation (any author country CN), share with a company affiliation, international share,
4+share with an arXiv version, mean/median c36 and Gini of c36 (complete windows only), mean NCS36, share in top 10 % FWCI.
5+Outputs results/bibliometrics/category_actors_yearly.csv and category_impact_yearly.csv (read-only DuckDB)."""
6+import argparse, os, duckdb
7+
8+
9+def main():
10+ ap = argparse.ArgumentParser(); ap.add_argument("--root", required=True); a = ap.parse_args(); R = a.root
11+ out = f"{R}/results/bibliometrics"; con = duckdb.connect(f"{R}/data/processed/wp44.duckdb", read_only=True); con.execute(f"SET threads TO 16; SET memory_limit='{os.environ.get('WP44_DUCKDB_MEM', '30GB')}';")
12+ P = f"read_parquet('{R}/data/processed/classification/primary/*.parquet')"; W = f"read_parquet('{R}/data/processed/corpus/work_citation_windows.parquet')"
13+ con.execute(f"""COPY (SELECT c.publication_year AS year, c.period, p.primary_category, count(*) n,
14+ round(avg(list_contains(c.countries, 'CN')::int), 4) share_cn, round(avg(list_contains(c.countries, 'US')::int), 4) share_us,
15+ round(avg(list_contains(c.institution_types, 'company')::int), 4) share_company, round(avg((c.n_countries >= 2)::int), 4) share_intl,
16+ round(avg((c.arxiv_id_linked IS NOT NULL)::int), 4) share_arxiv, round(avg(c.is_oa::int), 4) share_oa, round(avg(c.n_authors), 3) mean_authors
17+ FROM corpus_works c JOIN {P} p ON p.work_id = c.id WHERE c.eligible GROUP BY 1,2,3 ORDER BY 1,3) TO '{out}/category_actors_yearly.csv' (HEADER)""")
18+ con.execute(f"""COPY (WITH J AS (SELECT w.year, w.c36, w.c36_complete, w.ncs36, w.work_id, p.primary_category, c.is_top10 FROM {W} w JOIN {P} p ON p.work_id = w.work_id JOIN corpus_works c ON c.id = w.work_id),
19+ Rk AS (SELECT *, row_number() OVER (PARTITION BY year, primary_category ORDER BY c36, work_id) rk FROM J)
20+ SELECT year, primary_category, count(*) n, any_value(c36_complete) c36_complete, round(avg(c36), 3) mean_c36, quantile_cont(c36, 0.5) med_c36,
21+ round(sum((c36 = 0)::int) * 1.0 / count(*), 4) share_uncited_36, round(avg(ncs36), 3) mean_ncs36,
22+ round((2 * sum(c36 * rk) / (count(*) * sum(c36)) - (count(*) + 1.0) / count(*)), 4) gini_c36, round(avg(is_top10::int), 4) share_top10_fwci
23+ FROM Rk GROUP BY 1,2 HAVING count(*) >= 200 ORDER BY 1,2) TO '{out}/category_impact_yearly.csv' (HEADER)""")
24+ print(open(f"{out}/category_actors_yearly.csv").read()[:600])
25+
26+
27+if __name__ == "__main__":
28+ main()
modified figures/make_figures.py +1 −1
@@ -225,7 +225,7 @@ def fig_category_growth(R, out):
225 if not os.path.exists(p): 225 if not os.path.exists(p):
226 return 226 return
227 names, areas = _cat_names(R); d = pd.read_csv(p); d = d[(d.year >= 2018) & (d.year <= 2025)] 227 names, areas = _cat_names(R); d = pd.read_csv(p); d = d[(d.year >= 2018) & (d.year <= 2025)]
228− piv = d.pivot_table(index="year", columns="category_id", values="n_primary", aggfunc="sum", fill_value=0); base = piv.loc[2018] # aggfunc="sum" over the (year, period) rows (fixed 2026-11-04) 228+ piv = d.pivot_table(index="year", columns="category_id", values="n_primary", aggfunc="sum", fill_value=0); base = piv.loc[2018] # aggfunc="sum" over the (year, period) rows (fixed 2026-11-03)
229 idx = piv / base.replace(0, np.nan) * 100; cats = sorted(idx.columns, key=lambda c: -idx.loc[2025, c] if 2025 in idx.index else 0) 229 idx = piv / base.replace(0, np.nan) * 100; cats = sorted(idx.columns, key=lambda c: -idx.loc[2025, c] if 2025 in idx.index else 0)
230 fig, axes = plt.subplots(1, 2, figsize=(7.4, 3.6), sharey=False) 230 fig, axes = plt.subplots(1, 2, figsize=(7.4, 3.6), sharey=False)
231 for ax, sel, title in [(axes[0], cats[:8], "Eight fastest-growing categories"), (axes[1], cats[-8:], "Eight slowest-growing categories")]: 231 for ax, sel, title in [(axes[0], cats[:8], "Eight fastest-growing categories"), (axes[1], cats[-8:], "Eight slowest-growing categories")]:
modified manuscript/tables/TABLES.json +118 −55
@@ -103,7 +103,7 @@
103 "caption": "Fractional output of the fifteen largest countries by year.", 103 "caption": "Fractional output of the fifteen largest countries by year.",
104 "label": "tab:countries", 104 "label": "tab:countries",
105 "sources": [ 105 "sources": [
106− "./results/bibliometrics/countries_yearly.csv" 106+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/countries_yearly.csv"
107 ], 107 ],
108 "rows": 15 108 "rows": 15
109 }, 109 },
@@ -112,7 +112,7 @@
112 "caption": "Structural-change estimates by series: ITS level and slope changes at 2022-12 (HAC p-values), placebo p-values, Bai\u2013Perron (BIC) and PELT break dates.", 112 "caption": "Structural-change estimates by series: ITS level and slope changes at 2022-12 (HAC p-values), placebo p-values, Bai\u2013Perron (BIC) and PELT break dates.",
113 "label": "tab:breaks", 113 "label": "tab:breaks",
114 "sources": [ 114 "sources": [
115− "./results/econometrics/breaks.csv" 115+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv"
116 ], 116 ],
117 "rows": 41 117 "rows": 41
118 }, 118 },
@@ -121,7 +121,7 @@
121 "caption": "Out-of-time performance of the emerging-topic detectors at horizon 12 months (test cutoffs 2023-01\u20132024-10).", 121 "caption": "Out-of-time performance of the emerging-topic detectors at horizon 12 months (test cutoffs 2023-01\u20132024-10).",
122 "label": "tab:emerging_h12", 122 "label": "tab:emerging_h12",
123 "sources": [ 123 "sources": [
124− "./results/emerging/metrics_h12.json" 124+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h12.json"
125 ], 125 ],
126 "rows": 3 126 "rows": 3
127 }, 127 },
@@ -130,7 +130,7 @@
130 "caption": "Out-of-time performance of the emerging-topic detectors at horizon 24 months (test cutoffs 2023-01\u20132024-04).", 130 "caption": "Out-of-time performance of the emerging-topic detectors at horizon 24 months (test cutoffs 2023-01\u20132024-04).",
131 "label": "tab:emerging_h24", 131 "label": "tab:emerging_h24",
132 "sources": [ 132 "sources": [
133− "./results/emerging/metrics_h24.json" 133+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h24.json"
134 ], 134 ],
135 "rows": 3 135 "rows": 3
136 }, 136 },
@@ -139,7 +139,7 @@
139 "caption": "Preregistered known-emergence set: detection month, peak-growth month and lead time.", 139 "caption": "Preregistered known-emergence set: detection month, peak-growth month and lead time.",
140 "label": "tab:known_set", 140 "label": "tab:known_set",
141 "sources": [ 141 "sources": [
142− "./results/emerging/known_set_lead_times.csv" 142+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/known_set_lead_times.csv"
143 ], 143 ],
144 "rows": 12 144 "rows": 12
145 }, 145 },
@@ -148,8 +148,8 @@
148 "caption": "Team size and collaboration indicators by publication year (all eligible works; periods of the source file summed).", 148 "caption": "Team size and collaboration indicators by publication year (all eligible works; periods of the source file summed).",
149 "label": "tab:collaboration", 149 "label": "tab:collaboration",
150 "sources": [ 150 "sources": [
151− "./results/bibliometrics/team_size_yearly.csv", 151+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/team_size_yearly.csv",
152− "./results/bibliometrics/international_collab_yearly.csv" 152+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/international_collab_yearly.csv"
153 ], 153 ],
154 "rows": 9 154 "rows": 9
155 }, 155 },
@@ -158,7 +158,7 @@
158 "caption": "Share of typed affiliations by OpenAlex institution type and year (per cent of affiliation-work pairs).", 158 "caption": "Share of typed affiliations by OpenAlex institution type and year (per cent of affiliation-work pairs).",
159 "label": "tab:institution_types", 159 "label": "tab:institution_types",
160 "sources": [ 160 "sources": [
161− "./results/bibliometrics/institution_types_yearly.csv" 161+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/institution_types_yearly.csv"
162 ], 162 ],
163 "rows": 9 163 "rows": 9
164 }, 164 },
@@ -167,7 +167,7 @@
167 "caption": "Share of fractional output of the fifteen largest countries by year (per cent).", 167 "caption": "Share of fractional output of the fifteen largest countries by year (per cent).",
168 "label": "tab:country_shares", 168 "label": "tab:country_shares",
169 "sources": [ 169 "sources": [
170− "./results/bibliometrics/countries_yearly.csv" 170+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/countries_yearly.csv"
171 ], 171 ],
172 "rows": 15 172 "rows": 15
173 }, 173 },
@@ -176,7 +176,7 @@
176 "caption": "Concentration of output across countries by year, and the two largest producers under full counting.", 176 "caption": "Concentration of output across countries by year, and the two largest producers under full counting.",
177 "label": "tab:country_concentration", 177 "label": "tab:country_concentration",
178 "sources": [ 178 "sources": [
179− "./results/bibliometrics/countries_yearly.csv" 179+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/countries_yearly.csv"
180 ], 180 ],
181 "rows": 9 181 "rows": 9
182 }, 182 },
@@ -185,7 +185,7 @@
185 "caption": "The country co-authorship network by period: edges with at least fifty joint works.", 185 "caption": "The country co-authorship network by period: edges with at least fifty joint works.",
186 "label": "tab:country_network", 186 "label": "tab:country_network",
187 "sources": [ 187 "sources": [
188− "./results/bibliometrics/country_pairs_by_period.csv" 188+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/country_pairs_by_period.csv"
189 ], 189 ],
190 "rows": 2 190 "rows": 2
191 }, 191 },
@@ -194,7 +194,7 @@
194 "caption": "The twenty strongest country pairs by period (number of joint works).", 194 "caption": "The twenty strongest country pairs by period (number of joint works).",
195 "label": "tab:country_pairs", 195 "label": "tab:country_pairs",
196 "sources": [ 196 "sources": [
197− "./results/bibliometrics/country_pairs_by_period.csv" 197+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/country_pairs_by_period.csv"
198 ], 198 ],
199 "rows": 20 199 "rows": 20
200 }, 200 },
@@ -203,7 +203,7 @@
203 "caption": "Institutions with at least two hundred works per period: coverage and concentration (fractional counting).", 203 "caption": "Institutions with at least two hundred works per period: coverage and concentration (fractional counting).",
204 "label": "tab:institution_concentration", 204 "label": "tab:institution_concentration",
205 "sources": [ 205 "sources": [
206− "./results/bibliometrics/top_institutions_by_period.csv" 206+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_institutions_by_period.csv"
207 ], 207 ],
208 "rows": 3 208 "rows": 3
209 }, 209 },
@@ -212,7 +212,7 @@
212 "caption": "The twenty-five institutions with the largest fractional output in the post period, with their baseline output.", 212 "caption": "The twenty-five institutions with the largest fractional output in the post period, with their baseline output.",
213 "label": "tab:top_institutions", 213 "label": "tab:top_institutions",
214 "sources": [ 214 "sources": [
215− "./results/bibliometrics/top_institutions_by_period.csv" 215+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_institutions_by_period.csv"
216 ], 216 ],
217 "rows": 25 217 "rows": 25
218 }, 218 },
@@ -221,7 +221,7 @@
221 "caption": "Distribution of works over source types among the listed sources, by period (per cent).", 221 "caption": "Distribution of works over source types among the listed sources, by period (per cent).",
222 "label": "tab:source_types", 222 "label": "tab:source_types",
223 "sources": [ 223 "sources": [
224− "./results/bibliometrics/top_sources_by_period.csv" 224+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_sources_by_period.csv"
225 ], 225 ],
226 "rows": 2 226 "rows": 2
227 }, 227 },
@@ -230,7 +230,7 @@
230 "caption": "The twenty largest sources (venues) by period, with OpenAlex source type.", 230 "caption": "The twenty largest sources (venues) by period, with OpenAlex source type.",
231 "label": "tab:top_sources", 231 "label": "tab:top_sources",
232 "sources": [ 232 "sources": [
233− "./results/bibliometrics/top_sources_by_period.csv" 233+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_sources_by_period.csv"
234 ], 234 ],
235 "rows": 20 235 "rows": 20
236 }, 236 },
@@ -239,7 +239,7 @@
239 "caption": "Age of cited works by citing year (resolved references of eligible works; corpus version of the result file at generation time).", 239 "caption": "Age of cited works by citing year (resolved references of eligible works; corpus version of the result file at generation time).",
240 "label": "tab:cited_age", 240 "label": "tab:cited_age",
241 "sources": [ 241 "sources": [
242− "./results/graph/cited_age_yearly.csv" 242+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/cited_age_yearly.csv"
243 ], 243 ],
244 "rows": 9 244 "rows": 9
245 }, 245 },
@@ -248,7 +248,7 @@
248 "caption": "The ten most cited works of each publication year by 36-month window citations (c36), with total citations to date.", 248 "caption": "The ten most cited works of each publication year by 36-month window citations (c36), with total citations to date.",
249 "label": "tab:top_cited", 249 "label": "tab:top_cited",
250 "sources": [ 250 "sources": [
251− "./results/bibliometrics/top_cited_by_year.csv" 251+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_cited_by_year.csv"
252 ], 252 ],
253 "rows": 80 253 "rows": 80
254 }, 254 },
@@ -257,7 +257,7 @@
257 "caption": "DBLP records of 2018--2026 without an OpenAlex counterpart, by DBLP record type and year.", 257 "caption": "DBLP records of 2018--2026 without an OpenAlex counterpart, by DBLP record type and year.",
258 "label": "tab:dblp_only", 258 "label": "tab:dblp_only",
259 "sources": [ 259 "sources": [
260− "./results/bibliometrics/dblp_only_yearly.csv" 260+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/dblp_only_yearly.csv"
261 ], 261 ],
262 "rows": 9 262 "rows": 9
263 }, 263 },
@@ -266,7 +266,7 @@
266 "caption": "The OpenAlex \\texttt{created\\_date} relative to the publication date, by publication year.", 266 "caption": "The OpenAlex \\texttt{created\\_date} relative to the publication date, by publication year.",
267 "label": "tab:indexing_lag", 267 "label": "tab:indexing_lag",
268 "sources": [ 268 "sources": [
269− "./results/bibliometrics/indexing_lag.csv" 269+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/indexing_lag.csv"
270 ], 270 ],
271 "rows": 9 271 "rows": 9
272 }, 272 },
@@ -275,7 +275,7 @@
275 "caption": "Reference coverage by citing year: declared, resolved, internal and external references of eligible works.", 275 "caption": "Reference coverage by citing year: declared, resolved, internal and external references of eligible works.",
276 "label": "tab:reference_coverage", 276 "label": "tab:reference_coverage",
277 "sources": [ 277 "sources": [
278− "./results/graph/reference_coverage_yearly.csv" 278+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/reference_coverage_yearly.csv"
279 ], 279 ],
280 "rows": 9 280 "rows": 9
281 }, 281 },
@@ -284,16 +284,79 @@
284 "caption": "Destination of the resolved references of eligible works by citing year: inside the corpus, computer science outside it, and the seven largest other OpenAlex fields (per cent of resolved references).", 284 "caption": "Destination of the resolved references of eligible works by citing year: inside the corpus, computer science outside it, and the seven largest other OpenAlex fields (per cent of resolved references).",
285 "label": "tab:field_flows", 285 "label": "tab:field_flows",
286 "sources": [ 286 "sources": [
287− "./results/graph/field_flows_yearly.csv" 287+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/field_flows_yearly.csv"
288 ], 288 ],
289 "rows": 9 289 "rows": 9
290 }, 290 },
291+ {
292+ "file": "tab_ai_growth.tex",
293+ "caption": "Volume, compound annual growth and corpus share of the nine AI categories, baseline (2018--2022) and post (2023--2025) periods.",
294+ "label": "tab:ai_growth",
295+ "sources": [
296+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/cagr_by_category.csv"
297+ ],
298+ "rows": 9
299+ },
300+ {
301+ "file": "tab_ai_counts.tex",
302+ "caption": "Yearly multi-label counts of the nine AI categories (all period rows summed; 2026 incomplete).",
303+ "label": "tab:ai_counts",
304+ "sources": [
305+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv"
306+ ],
307+ "rows": 9
308+ },
309+ {
310+ "file": "tab_ai_breaks.tex",
311+ "caption": "Structural-change estimates for the nine AI categories: interrupted-time-series level and slope changes at 2022-12 (ChatGPT), level changes at the alternative event dates 2023-04 (GPT-4) and 2024-09 (o1), placebo p-values, and Bai--Perron and PELT break dates.",
312+ "label": "tab:ai_breaks",
313+ "sources": [
314+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv"
315+ ],
316+ "rows": 9
317+ },
318+ {
319+ "file": "tab_ai_flows.tex",
320+ "caption": "Destination of the in-corpus references of the nine AI categories in 2019, 2022 and 2025 (per cent of in-corpus references of the citing category).",
321+ "label": "tab:ai_flows",
322+ "sources": [
323+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_flows_yearly.csv"
324+ ],
325+ "rows": 27
326+ },
327+ {
328+ "file": "tab_ai_landmark_flags.tex",
329+ "caption": "Specialist screening of the 234 AI/ML landmark candidates: records by primary category and data-quality flag.",
330+ "label": "tab:ai_landmark_flags",
331+ "sources": [
332+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/agent_E/contributions.csv"
333+ ],
334+ "rows": 9
335+ },
336+ {
337+ "file": "tab_ai_contribution_types.tex",
338+ "caption": "Contribution type of the 234 AI/ML landmark candidates by primary category (specialist judgement from abstract or metadata).",
339+ "label": "tab:ai_contribution_types",
340+ "sources": [
341+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/agent_E/contributions.csv"
342+ ],
343+ "rows": 9
344+ },
345+ {
346+ "file": "tab_ai_landmarks.tex",
347+ "caption": "The 234 AI/ML landmark candidates with the specialist's contribution type and data-quality flag (titles in literature_synthesis/agent_E/contributions.csv).",
348+ "label": "tab:ai_landmarks",
349+ "sources": [
350+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/agent_E/contributions.csv"
351+ ],
352+ "rows": 234
353+ },
291 { 354 {
292 "file": "tab_security_counts.tex", 355 "file": "tab_security_counts.tex",
293 "caption": "Eligible works in the five security-and-privacy categories by publication year: works carrying the category (all) and works whose primary category it is (prim.).", 356 "caption": "Eligible works in the five security-and-privacy categories by publication year: works carrying the category (all) and works whose primary category it is (prim.).",
294 "label": "tab:security_counts", 357 "label": "tab:security_counts",
295 "sources": [ 358 "sources": [
296− "./results/graph/category_counts_yearly.csv" 359+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv"
297 ], 360 ],
298 "rows": 9 361 "rows": 9
299 }, 362 },
@@ -302,8 +365,8 @@
302 "caption": "Growth and structural-change summary for the five security-and-privacy categories (multi-label counts).", 365 "caption": "Growth and structural-change summary for the five security-and-privacy categories (multi-label counts).",
303 "label": "tab:security_summary", 366 "label": "tab:security_summary",
304 "sources": [ 367 "sources": [
305− "./results/graph/category_counts_yearly.csv", 368+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv",
306− "./results/econometrics/breaks.csv" 369+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv"
307 ], 370 ],
308 "rows": 5 371 "rows": 5
309 }, 372 },
@@ -312,8 +375,8 @@
312 "caption": "Co-classification of security-and-privacy works with the LLM, AI-safety and AI-agents categories, by publication year: number of works carrying both categories and share of the security category's works.", 375 "caption": "Co-classification of security-and-privacy works with the LLM, AI-safety and AI-agents categories, by publication year: number of works carrying both categories and share of the security category's works.",
313 "label": "tab:security_llm_cooccurrence", 376 "label": "tab:security_llm_cooccurrence",
314 "sources": [ 377 "sources": [
315− "./results/graph/category_cooccurrence_yearly.csv", 378+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_cooccurrence_yearly.csv",
316− "./results/graph/category_counts_yearly.csv" 379+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv"
317 ], 380 ],
318 "rows": 9 381 "rows": 9
319 }, 382 },
@@ -322,7 +385,7 @@
322 "caption": "Composition of the automatically selected landmark set of the security-and-privacy area, by category.", 385 "caption": "Composition of the automatically selected landmark set of the security-and-privacy area, by category.",
323 "label": "tab:security_landmarks", 386 "label": "tab:security_landmarks",
324 "sources": [ 387 "sources": [
325− "./literature_synthesis/landmarks/security_privacy_landmarks.csv" 388+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/landmarks/security_privacy_landmarks.csv"
326 ], 389 ],
327 "rows": 5 390 "rows": 5
328 }, 391 },
@@ -331,8 +394,8 @@
331 "caption": "The systems and infrastructure categories of Part VI: primary-category output, compound annual growth, and structural-change estimates of the monthly log-volume series.", 394 "caption": "The systems and infrastructure categories of Part VI: primary-category output, compound annual growth, and structural-change estimates of the monthly log-volume series.",
332 "label": "tab:systems_categories", 395 "label": "tab:systems_categories",
333 "sources": [ 396 "sources": [
334− "./results/bibliometrics/cagr_by_category.csv", 397+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/cagr_by_category.csv",
335− "./results/econometrics/breaks.csv" 398+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv"
336 ], 399 ],
337 "rows": 11 400 "rows": 11
338 }, 401 },
@@ -341,7 +404,7 @@
341 "caption": "The eligible corpus by protocol period: completeness, source membership and document type.", 404 "caption": "The eligible corpus by protocol period: completeness, source membership and document type.",
342 "label": "tab:p9-periods", 405 "label": "tab:p9-periods",
343 "sources": [ 406 "sources": [
344− "./results/bibliometrics/volume_monthly.csv" 407+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/volume_monthly.csv"
345 ], 408 ],
346 "rows": 13 409 "rows": 13
347 }, 410 },
@@ -350,8 +413,8 @@
350 "caption": "Classification coverage by publication year: eligible works, works with a primary category, and multi-label assignments.", 413 "caption": "Classification coverage by publication year: eligible works, works with a primary category, and multi-label assignments.",
351 "label": "tab:p9-classification", 414 "label": "tab:p9-classification",
352 "sources": [ 415 "sources": [
353− "./results/graph/category_counts_yearly.csv", 416+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv",
354− "./data/metadata/prisma_counts.json" 417+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/data/metadata/prisma_counts.json"
355 ], 418 ],
356 "rows": 9 419 "rows": 9
357 }, 420 },
@@ -360,7 +423,7 @@
360 "caption": "Share of primary-category assignments by top-level area and year (per cent), with the two protocol periods.", 423 "caption": "Share of primary-category assignments by top-level area and year (per cent), with the two protocol periods.",
361 "label": "tab:p9-area-shares", 424 "label": "tab:p9-area-shares",
362 "sources": [ 425 "sources": [
363− "./results/graph/category_counts_yearly.csv" 426+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv"
364 ], 427 ],
365 "rows": 11 428 "rows": 11
366 }, 429 },
@@ -369,8 +432,8 @@
369 "caption": "The forty categories by protocol period: primary-category output, composition shares, multi-label growth and compound annual growth rates.", 432 "caption": "The forty categories by protocol period: primary-category output, composition shares, multi-label growth and compound annual growth rates.",
370 "label": "tab:p9-category-periods", 433 "label": "tab:p9-category-periods",
371 "sources": [ 434 "sources": [
372− "./results/graph/category_counts_yearly.csv", 435+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv",
373− "./results/bibliometrics/cagr_by_category.csv" 436+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/cagr_by_category.csv"
374 ], 437 ],
375 "rows": 40 438 "rows": 40
376 }, 439 },
@@ -379,7 +442,7 @@
379 "caption": "Share of primary-category assignments held by the AI and machine-learning categories in 2018, 2022 and 2025 (per cent, bootstrap 95\\% intervals).", 442 "caption": "Share of primary-category assignments held by the AI and machine-learning categories in 2018, 2022 and 2025 (per cent, bootstrap 95\\% intervals).",
380 "label": "tab:p9-ai-shares", 443 "label": "tab:p9-ai-shares",
381 "sources": [ 444 "sources": [
382− "./results/bibliometrics/category_share_ci_yearly.csv" 445+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/category_share_ci_yearly.csv"
383 ], 446 ],
384 "rows": 10 447 "rows": 10
385 }, 448 },
@@ -388,7 +451,7 @@
388 "caption": "Interrupted time series of the total monthly series at the reference date and at four alternative event dates (log counts, month-of-year dummies, Newey--West standard errors).", 451 "caption": "Interrupted time series of the total monthly series at the reference date and at four alternative event dates (log counts, month-of-year dummies, Newey--West standard errors).",
389 "label": "tab:p9-its-total-events", 452 "label": "tab:p9-its-total-events",
390 "sources": [ 453 "sources": [
391− "./results/econometrics/its_results.json" 454+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
392 ], 455 ],
393 "rows": 5 456 "rows": 5
394 }, 457 },
@@ -397,7 +460,7 @@
397 "caption": "Robustness diagnostics of the structural-change analysis of the total monthly series.", 460 "caption": "Robustness diagnostics of the structural-change analysis of the total monthly series.",
398 "label": "tab:p9-total-robustness", 461 "label": "tab:p9-total-robustness",
399 "sources": [ 462 "sources": [
400− "./results/econometrics/its_results.json" 463+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
401 ], 464 ],
402 "rows": 12 465 "rows": 12
403 }, 466 },
@@ -406,7 +469,7 @@
406 "caption": "ITS slope change at December 2022 by category: pre-period slope, change, HAC and placebo $p$-values, and implied post-period slope (monthly log growth).", 469 "caption": "ITS slope change at December 2022 by category: pre-period slope, change, HAC and placebo $p$-values, and implied post-period slope (monthly log growth).",
407 "label": "tab:p9-its-slopes", 470 "label": "tab:p9-its-slopes",
408 "sources": [ 471 "sources": [
409− "./results/econometrics/its_results.json" 472+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
410 ], 473 ],
411 "rows": 40 474 "rows": 40
412 }, 475 },
@@ -415,7 +478,7 @@
415 "caption": "ITS level change at December 2022 by category: full-sample and donut estimates, HAC and placebo $p$-values, and the BOCPD change probability over December 2022 to March 2023.", 478 "caption": "ITS level change at December 2022 by category: full-sample and donut estimates, HAC and placebo $p$-values, and the BOCPD change probability over December 2022 to March 2023.",
416 "label": "tab:p9-its-levels", 479 "label": "tab:p9-its-levels",
417 "sources": [ 480 "sources": [
418− "./results/econometrics/its_results.json" 481+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
419 ], 482 ],
420 "rows": 40 483 "rows": 40
421 }, 484 },
@@ -424,7 +487,7 @@
424 "caption": "Data-driven break and changepoint dates by category, and whether any falls in the H3 window (October 2022 to June 2023).", 487 "caption": "Data-driven break and changepoint dates by category, and whether any falls in the H3 window (October 2022 to June 2023).",
425 "label": "tab:p9-unknown-breaks", 488 "label": "tab:p9-unknown-breaks",
426 "sources": [ 489 "sources": [
427− "./results/econometrics/its_results.json" 490+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
428 ], 491 ],
429 "rows": 40 492 "rows": 40
430 }, 493 },
@@ -433,7 +496,7 @@
433 "caption": "ITS level change (HAC $p$) by category at the five pre-specified event dates.", 496 "caption": "ITS level change (HAC $p$) by category at the five pre-specified event dates.",
434 "label": "tab:p9-altdates", 497 "label": "tab:p9-altdates",
435 "sources": [ 498 "sources": [
436− "./results/econometrics/its_results.json" 499+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
437 ], 500 ],
438 "rows": 40 501 "rows": 40
439 }, 502 },
@@ -442,7 +505,7 @@
442 "caption": "Hypothesis H3: break dates and ITS estimates for the eight categories named in the protocol.", 505 "caption": "Hypothesis H3: break dates and ITS estimates for the eight categories named in the protocol.",
443 "label": "tab:p9-h3", 506 "label": "tab:p9-h3",
444 "sources": [ 507 "sources": [
445− "./results/econometrics/its_results.json" 508+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json"
446 ], 509 ],
447 "rows": 8 510 "rows": 8
448 }, 511 },
@@ -451,7 +514,7 @@
451 "caption": "Cross-area share of internal references by citing area and citing year (per cent), with the aggregate share and its composition-adjusted counterpart.", 514 "caption": "Cross-area share of internal references by citing area and citing year (per cent), with the aggregate share and its composition-adjusted counterpart.",
452 "label": "tab:p9-cross-area", 515 "label": "tab:p9-cross-area",
453 "sources": [ 516 "sources": [
454− "./results/graph/cross_area_share_yearly.csv" 517+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/cross_area_share_yearly.csv"
455 ], 518 ],
456 "rows": 11 519 "rows": 11
457 }, 520 },
@@ -460,7 +523,7 @@
460 "caption": "Modularity of the yearly category citation-flow network at the category and area partitions, and the within-category self-share of citations.", 523 "caption": "Modularity of the yearly category citation-flow network at the category and area partitions, and the within-category self-share of citations.",
461 "label": "tab:p9-modularity", 524 "label": "tab:p9-modularity",
462 "sources": [ 525 "sources": [
463− "./results/graph/modularity_yearly.csv" 526+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/modularity_yearly.csv"
464 ], 527 ],
465 "rows": 9 528 "rows": 9
466 }, 529 },
@@ -469,7 +532,7 @@
469 "caption": "Change in the area citation-flow matrix between 2018--2022 and 2023--2026: post-period row share minus baseline row share (percentage points).", 532 "caption": "Change in the area citation-flow matrix between 2018--2022 and 2023--2026: post-period row share minus baseline row share (percentage points).",
470 "label": "tab:p9-area-flow-change", 533 "label": "tab:p9-area-flow-change",
471 "sources": [ 534 "sources": [
472− "./results/graph/area_flows_yearly.csv" 535+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/area_flows_yearly.csv"
473 ], 536 ],
474 "rows": 8 537 "rows": 8
475 }, 538 },
@@ -478,7 +541,7 @@
478 "caption": "The fifteen most frequent co-classification pairs (works carrying both categories in their label set), 2018--2022 and 2023--2026.", 541 "caption": "The fifteen most frequent co-classification pairs (works carrying both categories in their label set), 2018--2022 and 2023--2026.",
479 "label": "tab:p9-cooccurrence", 542 "label": "tab:p9-cooccurrence",
480 "sources": [ 543 "sources": [
481− "./results/graph/category_cooccurrence_yearly.csv" 544+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_cooccurrence_yearly.csv"
482 ], 545 ],
483 "rows": 15 546 "rows": 15
484 }, 547 },
@@ -487,7 +550,7 @@
487 "caption": "Categories most often co-assigned with Large Language Models, by span.", 550 "caption": "Categories most often co-assigned with Large Language Models, by span.",
488 "label": "tab:p9-llm-partners", 551 "label": "tab:p9-llm-partners",
489 "sources": [ 552 "sources": [
490− "./results/graph/category_cooccurrence_yearly.csv" 553+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_cooccurrence_yearly.csv"
491 ], 554 ],
492 "rows": 15 555 "rows": 15
493 }, 556 },
@@ -496,7 +559,7 @@
496 "caption": "Openness and metadata indicators by publication year: arXiv linkage, preprint share, open access, abstract availability and document type (per cent of eligible works).", 559 "caption": "Openness and metadata indicators by publication year: arXiv linkage, preprint share, open access, abstract availability and document type (per cent of eligible works).",
497 "label": "tab:p9-openness", 560 "label": "tab:p9-openness",
498 "sources": [ 561 "sources": [
499− "./results/bibliometrics/volume_yearly.csv" 562+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/volume_yearly.csv"
500 ], 563 ],
501 "rows": 9 564 "rows": 9
502 }, 565 },
@@ -505,7 +568,7 @@
505 "caption": "Composition of the landmark set by area and by publication year.", 568 "caption": "Composition of the landmark set by area and by publication year.",
506 "label": "tab:p9-landmarks", 569 "label": "tab:p9-landmarks",
507 "sources": [ 570 "sources": [
508− "./literature_synthesis/landmarks/selection_report.json" 571+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/landmarks/selection_report.json"
509 ], 572 ],
510 "rows": 17 573 "rows": 17
511 }, 574 },
@@ -514,8 +577,8 @@
514 "caption": "Out-of-time evaluation design of the emerging-topic detector: cutoffs, panel sizes and positive rates.", 577 "caption": "Out-of-time evaluation design of the emerging-topic detector: cutoffs, panel sizes and positive rates.",
515 "label": "tab:p9-emerging-setup", 578 "label": "tab:p9-emerging-setup",
516 "sources": [ 579 "sources": [
517− "./results/emerging/metrics_h12.json", 580+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h12.json",
518− "./results/emerging/metrics_h24.json" 581+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h24.json"
519 ], 582 ],
520 "rows": 2 583 "rows": 2
521 }, 584 },
@@ -524,8 +587,8 @@
524 "caption": "AUC of the three detectors at each test cutoff, by horizon.", 587 "caption": "AUC of the three detectors at each test cutoff, by horizon.",
525 "label": "tab:p9-emerging-cutoffs", 588 "label": "tab:p9-emerging-cutoffs",
526 "sources": [ 589 "sources": [
527− "./results/emerging/metrics_h12.json", 590+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h12.json",
528− "./results/emerging/metrics_h24.json" 591+ "/Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h24.json"
529 ], 592 ],
530 "rows": 8 593 "rows": 8
531 } 594 }
added manuscript/tables/tab_ai_breaks.tex +23 −0
@@ -0,0 +1,23 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv
2+\begin{longtable}{lrrrrrrrrrrrr}
3+\caption{Structural-change estimates for the nine AI categories: interrupted-time-series level and slope changes at 2022-12 (ChatGPT), level changes at the alternative event dates 2023-04 (GPT-4) and 2024-09 (o1), placebo p-values, and Bai--Perron and PELT break dates.}\label{tab:ai_breaks}\\
4+\toprule
5+\textbf{Category} & \textbf{Level 2022-12} & \textbf{p} & \textbf{Slope 2022-12} & \textbf{p } & \textbf{Level 2023-04} & \textbf{p } & \textbf{Level 2024-09} & \textbf{p } & \textbf{Placebo p (level)} & \textbf{Placebo p (slope)} & \textbf{Bai--Perron (BIC)} & \textbf{PELT} \\
6+\midrule
7+\endfirsthead
8+\toprule
9+\textbf{Category} & \textbf{Level 2022-12} & \textbf{p} & \textbf{Slope 2022-12} & \textbf{p } & \textbf{Level 2023-04} & \textbf{p } & \textbf{Level 2024-09} & \textbf{p } & \textbf{Placebo p (level)} & \textbf{Placebo p (slope)} & \textbf{Bai--Perron (BIC)} & \textbf{PELT} \\
10+\midrule
11+\endhead
12+Large language models & 0.106 & 0.075 & 0.034 & 0.000 & 0.236 & 0.002 & 0.457 & 0.001 & 0.938 & 0.200 & 2022-07-01 & 2023-10 \\
13+Foundation models & -0.075 & 0.340 & 0.006 & 0.021 & 0.023 & 0.756 & 0.084 & 0.173 & 0.692 & 0.246 & 2021-10-01 & 2020-02; 2023-10 \\
14+Multimodal AI & 0.025 & 0.600 & 0.024 & 0.000 & 0.150 & 0.002 & 0.330 & 0.001 & 0.969 & 0.200 & 2022-04-01 & 2020-12; 2023-12 \\
15+Generative AI & 0.036 & 0.556 & 0.019 & 0.000 & 0.150 & 0.017 & 0.324 & 0.000 & 0.969 & 0.308 & 2021-11-01 & 2020-12; 2023-12 \\
16+AI agents & -0.210 & 0.028 & 0.040 & 0.000 & -0.151 & 0.090 & 0.226 & 0.003 & 0.631 & 0.400 & 2022-07-01; 2024-07-01 & 2023-12; 2025-08 \\
17+Reinforcement learning & -0.164 & 0.013 & 0.012 & 0.000 & -0.158 & 0.006 & 0.054 & 0.071 & 0.615 & 0.385 & 2021-07-01; 2024-12-01 & 2021-01; 2024-12 \\
18+ML theory & -0.080 & 0.090 & 0.009 & 0.000 & -0.047 & 0.296 & 0.081 & 0.026 & 0.662 & 0.215 & 2020-05-01; 2022-07-01 & 2020-02; 2023-12 \\
19+DL architectures & -0.071 & 0.126 & -0.002 & 0.136 & -0.051 & 0.260 & 0.008 & 0.792 & 0.400 & 0.708 & 2021-07-01 & 2020-02; 2023-12 \\
20+AI safety and alignment & -0.206 & 0.002 & 0.016 & 0.000 & -0.080 & 0.244 & 0.150 & 0.010 & 0.446 & 0.308 & 2021-06-01; 2022-09-01 & 2020-02; 2024-02 \\
21+\bottomrule
22+\multicolumn{13}{p{0.95\linewidth}}{\footnotesize Monthly log counts, seasonally adjusted; HAC p-values (Part IV, Section~\ref{sec:m-its}). Level changes are in log points of the monthly count; slope changes in log points per month. Placebo p-values are the share of placebo event dates (Section~\ref{sec:m-placebo}) whose t-statistic exceeds the true one. Bai--Perron dates: breaks selected by BIC with trimming h=15 months; PELT: penalised exact linear-time change points. The categories are multi-label, so the series overlap.}\\
23+\end{longtable}
added manuscript/tables/tab_ai_contribution_types.tex +20 −0
@@ -0,0 +1,20 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/agent_E/contributions.csv
2+\begin{table}[htbp]\centering\caption{Contribution type of the 234 AI/ML landmark candidates by primary category (specialist judgement from abstract or metadata).}\label{tab:ai_contribution_types}
3+\begin{threeparttable}\small
4+\begin{tabular}{lrrrrrrr}
5+\toprule
6+\textbf{Primary category} & \textbf{benchmark} & \textbf{empirical} & \textbf{engineering} & \textbf{method} & \textbf{survey} & \textbf{theory} & \textbf{total} \\
7+\midrule
8+Large language models & 1 & 8 & 6 & 2 & 12 & 4 & 33 \\
9+Foundation models & 1 & 0 & 1 & 16 & 6 & 1 & 25 \\
10+Multimodal AI & 1 & 2 & 0 & 16 & 6 & 0 & 25 \\
11+Generative AI & 0 & 9 & 1 & 5 & 8 & 2 & 25 \\
12+AI agents & 1 & 0 & 4 & 0 & 13 & 7 & 25 \\
13+Reinforcement learning & 0 & 5 & 2 & 11 & 3 & 4 & 25 \\
14+ML theory & 0 & 0 & 1 & 4 & 11 & 9 & 25 \\
15+DL architectures & 0 & 2 & 1 & 11 & 11 & 0 & 25 \\
16+AI safety and alignment & 0 & 7 & 1 & 2 & 14 & 2 & 26 \\
17+\bottomrule
18+\end{tabular}
19+\begin{tablenotes}\footnotesize\item Types follow CLAUDE.md~\S1 item 6: theory, method, empirical, benchmark, engineering, survey. Judged from the abstract for 190 records and from title/venue/type for 44 records without abstract (column judged\_from of the source file).\end{tablenotes}
20+\end{threeparttable}\end{table}
added manuscript/tables/tab_ai_counts.tex +20 −0
@@ -0,0 +1,20 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv
2+\begin{table}[htbp]\centering\caption{Yearly multi-label counts of the nine AI categories (all period rows summed; 2026 incomplete).}\label{tab:ai_counts}
3+\begin{threeparttable}\small
4+\begin{tabular}{lrrrrrrrrr}
5+\toprule
6+\textbf{Category} & \textbf{2018} & \textbf{2019} & \textbf{2020} & \textbf{2021} & \textbf{2022} & \textbf{2023} & \textbf{2024} & \textbf{2025} & \textbf{2026} \\
7+\midrule
8+Large language models & 15{,}397 & 18{,}253 & 21{,}706 & 24{,}153 & 25{,}206 & 40{,}491 & 69{,}900 & 116{,}242 & 132{,}202 \\
9+Foundation models & 4{,}486 & 7{,}802 & 11{,}029 & 14{,}509 & 16{,}948 & 25{,}052 & 36{,}652 & 52{,}535 & 56{,}874 \\
10+Multimodal AI & 3{,}989 & 5{,}553 & 6{,}666 & 8{,}561 & 10{,}369 & 15{,}746 & 24{,}979 & 42{,}707 & 47{,}737 \\
11+Generative AI & 5{,}118 & 8{,}354 & 10{,}228 & 12{,}580 & 13{,}703 & 21{,}583 & 36{,}186 & 56{,}768 & 55{,}673 \\
12+AI agents & 10{,}042 & 10{,}745 & 10{,}642 & 10{,}593 & 10{,}324 & 12{,}255 & 16{,}189 & 29{,}517 & 45{,}112 \\
13+Reinforcement learning & 8{,}157 & 11{,}558 & 14{,}307 & 16{,}674 & 18{,}491 & 22{,}678 & 28{,}572 & 43{,}949 & 44{,}996 \\
14+ML theory & 28{,}277 & 37{,}723 & 43{,}455 & 47{,}121 & 48{,}940 & 58{,}770 & 70{,}964 & 91{,}772 & 84{,}417 \\
15+DL architectures & 43{,}313 & 62{,}096 & 74{,}987 & 90{,}347 & 99{,}122 & 118{,}793 & 142{,}181 & 174{,}461 & 130{,}581 \\
16+AI safety and alignment & 3{,}309 & 5{,}385 & 7{,}749 & 9{,}909 & 11{,}593 & 16{,}606 & 26{,}098 & 44{,}888 & 51{,}629 \\
17+\bottomrule
18+\end{tabular}
19+\begin{tablenotes}\footnotesize\item The source file splits each year into period rows (baseline / pre / post relative to 30 November 2022); they are summed here. 2026 covers publications dated up to the snapshot of 14 October 2026.\end{tablenotes}
20+\end{threeparttable}\end{table}
added manuscript/tables/tab_ai_flows.tex +41 −0
@@ -0,0 +1,41 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_flows_yearly.csv
2+\begin{longtable}{lrrrrrrrrr}
3+\caption{Destination of the in-corpus references of the nine AI categories in 2019, 2022 and 2025 (per cent of in-corpus references of the citing category).}\label{tab:ai_flows}\\
4+\toprule
5+\textbf{Category} & \textbf{Year} & \textbf{Self} & \textbf{LLM} & \textbf{NLP} & \textbf{Vision} & \textbf{HCI} & \textbf{Robotics} & \textbf{Softw. eng.} & \textbf{AI safety} \\
6+\midrule
7+\endfirsthead
8+\toprule
9+\textbf{Category} & \textbf{Year} & \textbf{Self} & \textbf{LLM} & \textbf{NLP} & \textbf{Vision} & \textbf{HCI} & \textbf{Robotics} & \textbf{Softw. eng.} & \textbf{AI safety} \\
10+\midrule
11+\endhead
12+Large language models & 2019 & 3.8 & 3.8 & 69.6 & 4.7 & 0.7 & 0.2 & 0.6 & 0.2 \\
13+Large language models & 2022 & 7.5 & 7.5 & 68.6 & 4.5 & 0.9 & 0.2 & 1.2 & 0.4 \\
14+Large language models & 2025 & 28.4 & 28.4 & 34.2 & 4.5 & 4.3 & 1.0 & 3.4 & 2.4 \\
15+Foundation models & 2019 & 1.2 & 0.0 & 18.7 & 31.3 & 0.0 & 2.0 & 0.0 & 0.0 \\
16+Foundation models & 2022 & 2.9 & 1.5 & 10.5 & 37.4 & 2.9 & 2.1 & 0.3 & 0.4 \\
17+Foundation models & 2025 & 3.6 & 4.1 & 6.7 & 43.8 & 1.9 & 1.0 & 0.6 & 0.6 \\
18+Multimodal AI & 2019 & 12.9 & 0.2 & 15.7 & 51.1 & 1.0 & 1.5 & 0.2 & 0.3 \\
19+Multimodal AI & 2022 & 15.2 & 1.0 & 18.5 & 50.3 & 0.7 & 0.9 & 0.2 & 0.1 \\
20+Multimodal AI & 2025 & 14.3 & 3.7 & 16.1 & 46.2 & 1.5 & 1.7 & 0.3 & 0.3 \\
21+Generative AI & 2019 & 13.3 & 0.5 & 3.0 & 53.6 & 0.5 & 0.6 & 0.3 & 0.3 \\
22+Generative AI & 2022 & 10.7 & 0.2 & 2.1 & 57.9 & 1.3 & 0.6 & 0.4 & 0.3 \\
23+Generative AI & 2025 & 13.6 & 13.7 & 3.5 & 33.8 & 6.6 & 1.5 & 1.3 & 3.7 \\
24+AI agents & 2019 & 45.5 & 0.1 & 2.7 & 1.7 & 1.8 & 7.5 & 2.3 & 0.9 \\
25+AI agents & 2022 & 39.3 & 0.2 & 2.1 & 1.6 & 2.9 & 8.8 & 1.5 & 0.9 \\
26+AI agents & 2025 & 30.7 & 4.2 & 3.8 & 1.9 & 4.3 & 12.0 & 1.3 & 1.5 \\
27+Reinforcement learning & 2019 & 47.2 & 0.0 & 2.2 & 5.0 & 0.8 & 14.4 & 0.4 & 0.3 \\
28+Reinforcement learning & 2022 & 43.5 & 0.1 & 1.2 & 5.0 & 1.0 & 19.2 & 0.6 & 0.3 \\
29+Reinforcement learning & 2025 & 39.0 & 0.5 & 0.9 & 4.4 & 0.8 & 23.1 & 0.5 & 0.3 \\
30+ML theory & 2019 & 44.1 & 0.2 & 3.0 & 13.7 & 0.8 & 1.1 & 1.2 & 1.1 \\
31+ML theory & 2022 & 40.8 & 0.2 & 3.6 & 15.5 & 1.3 & 1.0 & 1.5 & 1.6 \\
32+ML theory & 2025 & 38.5 & 0.7 & 2.8 & 15.4 & 1.3 & 1.1 & 1.6 & 1.7 \\
33+DL architectures & 2019 & 36.9 & 0.2 & 5.1 & 21.8 & 0.7 & 1.5 & 1.2 & 0.2 \\
34+DL architectures & 2022 & 35.1 & 0.4 & 5.4 & 22.1 & 1.0 & 1.4 & 1.0 & 0.3 \\
35+DL architectures & 2025 & 35.3 & 1.0 & 4.3 & 21.8 & 1.0 & 1.4 & 0.7 & 0.3 \\
36+AI safety and alignment & 2019 & 27.1 & 0.1 & 7.0 & 6.6 & 9.1 & 3.2 & 4.1 & 27.1 \\
37+AI safety and alignment & 2022 & 34.4 & 0.8 & 8.2 & 7.3 & 10.0 & 2.3 & 3.6 & 34.4 \\
38+AI safety and alignment & 2025 & 28.9 & 10.0 & 8.1 & 5.2 & 11.5 & 2.5 & 2.7 & 28.9 \\
39+\bottomrule
40+\multicolumn{10}{p{0.95\linewidth}}{\footnotesize Row shares of category-to-category citation flows (Part IV, eq.~\ref{eq:flow}); references outside the eligible corpus are excluded. `Self' is the share going to the citing category itself (for LLM, the Self and LLM columns coincide). Categories are multi-label, so a reference can count under several destination categories.}\\
41+\end{longtable}
added manuscript/tables/tab_ai_growth.tex +20 −0
@@ -0,0 +1,20 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/cagr_by_category.csv
2+\begin{table}[htbp]\centering\caption{Volume, compound annual growth and corpus share of the nine AI categories, baseline (2018--2022) and post (2023--2025) periods.}\label{tab:ai_growth}
3+\begin{threeparttable}\small
4+\begin{tabular}{lrrrrrrrrr}
5+\toprule
6+\textbf{Category} & \textbf{2018} & \textbf{2022} & \textbf{2023} & \textbf{2025} & \textbf{CAGR 2018--22} & \textbf{CAGR 2023--25} & \textbf{Share 2018 (\%)} & \textbf{Share 2022 (\%)} & \textbf{Share 2025 (\%)} \\
7+\midrule
8+Large language models & 1{,}635 & 3{,}266 & 10{,}389 & 45{,}158 & 0.189 & 1.085 & 0.33 & 0.58 & 5.06 \\
9+Foundation models & 46 & 350 & 571 & 1{,}854 & 0.661 & 0.802 & 0.01 & 0.06 & 0.21 \\
10+Multimodal AI & 470 & 1{,}278 & 1{,}932 & 6{,}749 & 0.284 & 0.869 & 0.09 & 0.23 & 0.76 \\
11+Generative AI & 766 & 2{,}160 & 3{,}646 & 12{,}743 & 0.296 & 0.870 & 0.15 & 0.39 & 1.43 \\
12+AI agents & 5{,}436 & 4{,}521 & 4{,}991 & 8{,}819 & -0.045 & 0.329 & 1.09 & 0.81 & 0.99 \\
13+Reinforcement learning & 3{,}296 & 7{,}017 & 8{,}422 & 13{,}210 & 0.208 & 0.252 & 0.66 & 1.25 & 1.48 \\
14+ML theory & 13{,}534 & 20{,}131 & 22{,}893 & 32{,}798 & 0.104 & 0.197 & 2.71 & 3.59 & 3.67 \\
15+DL architectures & 11{,}992 & 22{,}270 & 25{,}607 & 34{,}141 & 0.167 & 0.155 & 2.40 & 3.97 & 3.82 \\
16+AI safety and alignment & 1{,}052 & 2{,}954 & 4{,}015 & 10{,}885 & 0.294 & 0.647 & 0.21 & 0.53 & 1.22 \\
17+\bottomrule
18+\end{tabular}
19+\begin{tablenotes}\footnotesize\item Counts are multi-label (a work counts in every category it is assigned to), restricted to the eligible mapping corpus. CAGR as in Part IV, eq.~\ref{eq:cagr}; the post-period CAGR uses 2023 and 2025 end points. Shares are over all eligible works of the year.\end{tablenotes}
20+\end{threeparttable}\end{table}
added manuscript/tables/tab_ai_landmark_flags.tex +20 −0
@@ -0,0 +1,20 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/agent_E/contributions.csv
2+\begin{table}[htbp]\centering\caption{Specialist screening of the 234 AI/ML landmark candidates: records by primary category and data-quality flag.}\label{tab:ai_landmark_flags}
3+\begin{threeparttable}\small
4+\begin{tabular}{lrrrrrrrr}
5+\toprule
6+\textbf{Primary category} & \textbf{abstract\_mismatch} & \textbf{identifier\_mismatch} & \textbf{no\_abstract} & \textbf{normalisation\_artefact} & \textbf{off\_scope} & \textbf{ok} & \textbf{year\_implausible} & \textbf{total} \\
7+\midrule
8+Large language models & 1 & 7 & 4 & 2 & 0 & 19 & 0 & 33 \\
9+Foundation models & 0 & 0 & 6 & 0 & 0 & 19 & 0 & 25 \\
10+Multimodal AI & 0 & 0 & 5 & 0 & 0 & 20 & 0 & 25 \\
11+Generative AI & 0 & 0 & 5 & 1 & 1 & 18 & 0 & 25 \\
12+AI agents & 0 & 0 & 2 & 1 & 4 & 18 & 0 & 25 \\
13+Reinforcement learning & 0 & 1 & 4 & 1 & 4 & 12 & 3 & 25 \\
14+ML theory & 1 & 1 & 6 & 2 & 0 & 11 & 4 & 25 \\
15+DL architectures & 0 & 3 & 4 & 0 & 0 & 14 & 4 & 25 \\
16+AI safety and alignment & 1 & 1 & 4 & 0 & 1 & 19 & 0 & 26 \\
17+\bottomrule
18+\end{tabular}
19+\begin{tablenotes}\footnotesize\item Flags assigned by Agent E from title, venue, identifiers, citation counts and (where available) the abstract: identifier\_mismatch = title/abstract do not correspond to the arXiv identifier or citation count attached to the record; abstract\_mismatch = stored abstract belongs to another work; year\_implausible = OpenAlex year contradicts the first publication; normalisation\_artefact = very small c36 with very large NCS36; off\_scope = not an AI contribution; no\_abstract = metadata only. `ok' records were judged from the abstract.\end{tablenotes}
20+\end{threeparttable}\end{table}
added manuscript/tables/tab_ai_landmarks.tex +248 −0
@@ -0,0 +1,248 @@
1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/agent_E/contributions.csv
2+\begin{longtable}{lrrrrrrr}
3+\caption{The 234 AI/ML landmark candidates with the specialist's contribution type and data-quality flag (titles in literature_synthesis/agent_E/contributions.csv).}\label{tab:ai_landmarks}\\
4+\toprule
5+\textbf{OpenAlex id} & \textbf{Year} & \textbf{Type} & \textbf{Category} & \textbf{c36} & \textbf{NCS36} & \textbf{Contribution} & \textbf{Flag} \\
6+\midrule
7+\endfirsthead
8+\toprule
9+\textbf{OpenAlex id} & \textbf{Year} & \textbf{Type} & \textbf{Category} & \textbf{c36} & \textbf{NCS36} & \textbf{Contribution} & \textbf{Flag} \\
10+\midrule
11+\endhead
12+7172491899 & 2026 & article & Large language models & 3{,}026 & 3367.3 & engineering & identifier\_mismatch \\
13+7154011589 & 2026 & article & Large language models & 2{,}847 & 3168.1 & empirical & identifier\_mismatch \\
14+4362515116 & 2026 & article & Large language models & 1{,}549 & 1723.7 & survey & ok \\
15+4288253152 & 2024 & dissertation & Large language models & 364 & 3102.1 & engineering & normalisation\_artefact \\
16+4322718191 & 2023 & preprint & Large language models & 3{,}629 & 1206.5 & engineering & ok \\
17+4323655724 & 2023 & article & Large language models & 4{,}227 & 990.7 & survey & ok \\
18+4367369313 & 2026 & article & Large language models & 261 & 290.4 & survey & ok \\
19+4226278401 & 2022 & preprint & Large language models & 3{,}077 & 894.2 & method & identifier\_mismatch \\
20+4396531244 & 2024 & dissertation & Large language models & 20 & 778.8 & engineering & normalisation\_artefact \\
21+4221143046 & 2022 & preprint & Large language models & 2{,}465 & 716.4 & theory & identifier\_mismatch \\
22+4327810158 & 2023 & preprint & Large language models & 1{,}968 & 654.3 & engineering & abstract\_mismatch \\
23+7135096569 & 2026 & article & Large language models & 183 & 203.6 & empirical & ok \\
24+4414281281 & 2025 & article & Large language models & 1{,}099 & 374.4 & method & ok \\
25+4389437528 & 2023 & article & Large language models & 2{,}361 & 553.4 & survey & ok \\
26+4292779060 & 2023 & article & Large language models & 13{,}333 & 534.7 & theory & identifier\_mismatch \\
27+4360836968 & 2023 & preprint & Large language models & 1{,}420 & 472.1 & empirical & ok \\
28+4406800520 & 2025 & article & Large language models & 914 & 311.4 & survey & no\_abstract \\
29+4385223296 & 2023 & book-chapter & Large language models & 612 & 421.3 & survey & no\_abstract \\
30+4287689466 & 2025 & conference-paper & Large language models & 722 & 548.6 & empirical & identifier\_mismatch \\
31+4366420437 & 2023 & article & Large language models & 1{,}541 & 361.2 & survey & ok \\
32+4317910584 & 2023 & article & Large language models & 1{,}530 & 358.6 & survey & ok \\
33+4313564799 & 2023 & preprint & Large language models & 811 & 335.9 & empirical & no\_abstract \\
34+4382618722 & 2026 & article & Large language models & 94 & 98.9 & engineering & ok \\
35+4384464487 & 2023 & article & Large language models & 1{,}398 & 327.7 & empirical & ok \\
36+4392773210 & 2024 & article & Large language models & 1{,}085 & 326.7 & survey & ok \\
37+4387355843 & 2026 & book-chapter & Foundation models & 197 & 188.6 & method & ok \\
38+4392947521 & 2024 & article & Foundation models & 1{,}505 & 186.9 & method & no\_abstract \\
39+7124890304 & 2026 & article & Foundation models & 42 & 46.5 & method & no\_abstract \\
40+4410551352 & 2025 & article & Foundation models & 249 & 69.7 & method & ok \\
41+4322832366 & 2026 & conference-paper & Foundation models & 35 & 126.3 & method & ok \\
42+7133233472 & 2023 & conference-paper & Foundation models & 532 & 122.7 & method & ok \\
43+2990704537 & 2019 & article & Foundation models & 668 & 101.6 & benchmark & ok \\
44+7125271461 & 2026 & article & Foundation models & 25 & 70.6 & engineering & ok \\
45+7119541039 & 2026 & article & Foundation models & 24 & 26.7 & theory & ok \\
46+4393138529 & 2024 & preprint & Foundation models & 95 & 82.9 & survey & ok \\
47+4390690043 & 2026 & article & Foundation models & 22 & 10.0 & survey & ok \\
48+7128029160 & 2026 & article & Foundation models & 20 & 22.3 & method & ok \\
49+4402220065 & 2024 & article & Foundation models & 562 & 69.8 & method & no\_abstract \\
50+7118170538 & 2026 & article & Foundation models & 17 & 17.9 & survey & ok \\
51+4411932262 & 2025 & article & Foundation models & 110 & 26.9 & method & ok \\
52+4411099958 & 2025 & article & Foundation models & 108 & 18.4 & method & ok \\
53+4361229539 & 2023 & preprint & Foundation models & 161 & 53.5 & method & ok \\
54+4407135816 & 2025 & article & Foundation models & 103 & 6.9 & survey & ok \\
55+4385948838 & 2023 & article & Foundation models & 471 & 49.1 & method & no\_abstract \\
56+4406643700 & 2025 & article & Foundation models & 91 & 19.1 & survey & ok \\
57+4406472599 & 2025 & article & Foundation models & 86 & 29.3 & survey & no\_abstract \\
58+4393153153 & 2024 & conference-paper & Foundation models & 113 & 41.8 & method & ok \\
59+4402727669 & 2024 & conference-paper & Foundation models & 184 & 38.6 & method & ok \\
60+4401397240 & 2024 & article & Foundation models & 325 & 38.5 & method & no\_abstract \\
61+4385285835 & 2023 & preprint & Foundation models & 107 & 35.6 & method & ok \\
62+4318718936 & 2023 & preprint & Multimodal AI & 875 & 273.1 & method & ok \\
63+4393148714 & 2024 & conference-paper & Multimodal AI & 923 & 193.7 & method & ok \\
64+4396599905 & 2026 & article & Multimodal AI & 48 & 128.7 & survey & ok \\
65+4396952261 & 2024 & book-chapter & Multimodal AI & 189 & 147.2 & survey & no\_abstract \\
66+7133250524 & 2022 & conference-paper & Multimodal AI & 828 & 111.9 & benchmark & no\_abstract \\
67+4392947532 & 2024 & article & Multimodal AI & 938 & 111.1 & method & no\_abstract \\
68+7138474371 & 2026 & conference-paper & Multimodal AI & 31 & 159.4 & method & ok \\
69+7127319731 & 2026 & article & Multimodal AI & 29 & 77.7 & method & ok \\
70+2619383789 & 2018 & article & Multimodal AI & 587 & 103.1 & survey & ok \\
71+7148475777 & 2026 & conference-paper & Multimodal AI & 26 & 73.2 & empirical & ok \\
72+7160627408 & 2026 & conference-paper & Multimodal AI & 23 & 193.4 & method & ok \\
73+2995460200 & 2020 & conference-paper & Multimodal AI & 670 & 81.1 & method & ok \\
74+4415913530 & 2025 & article & Multimodal AI & 157 & 53.5 & method & ok \\
75+4289785095 & 2022 & preprint & Multimodal AI & 366 & 79.4 & method & ok \\
76+7125233009 & 2026 & article & Multimodal AI & 22 & 24.5 & method & ok \\
77+7143512276 & 2026 & article & Multimodal AI & 21 & 23.4 & survey & ok \\
78+4413447271 & 2025 & article & Multimodal AI & 142 & 46.4 & method & ok \\
79+7131830459 & 2026 & article & Multimodal AI & 20 & 142.1 & survey & ok \\
80+7206206818 & 2026 & article & Multimodal AI & 18 & 48.2 & method & ok \\
81+7133220561 & 2023 & conference-paper & Multimodal AI & 454 & 63.4 & method & no\_abstract \\
82+7164501397 & 2026 & article & Multimodal AI & 17 & 45.6 & empirical & ok \\
83+7165561001 & 2026 & conference-paper & Multimodal AI & 17 & 143.0 & method & no\_abstract \\
84+7138470740 & 2026 & conference-paper & Multimodal AI & 17 & 130.5 & method & ok \\
85+7138363754 & 2026 & conference-paper & Multimodal AI & 17 & 130.5 & method & ok \\
86+4410882873 & 2025 & article & Multimodal AI & 119 & 7.9 & survey & ok \\
87+4400228804 & 2024 & dissertation & Generative AI & 2 & 540.0 & empirical & normalisation\_artefact \\
88+7124262017 & 2026 & book & Generative AI & 87 & 783.7 & survey & ok \\
89+4406472463 & 2025 & article & Generative AI & 498 & 77.1 & method & ok \\
90+7123342962 & 2026 & article & Generative AI & 66 & 176.9 & method & ok \\
91+4383312437 & 2023 & article & Generative AI & 988 & 231.6 & empirical & ok \\
92+7133185348 & 2022 & conference-paper & Generative AI & 1{,}706 & 230.5 & method & no\_abstract \\
93+3195296536 & 2020 & dissertation & Generative AI & 48 & 221.7 & method & no\_abstract \\
94+4322617812 & 2023 & article & Generative AI & 873 & 204.6 & theory & ok \\
95+3096831136 & 2020 & article & Generative AI & 1{,}473 & 198.6 & survey & ok \\
96+4400770292 & 2024 & preprint & Generative AI & 198 & 174.2 & empirical & no\_abstract \\
97+7127274868 & 2026 & article & Generative AI & 48 & 341.0 & empirical & ok \\
98+7124974472 & 2026 & article & Generative AI & 42 & 46.7 & empirical & ok \\
99+7118187240 & 2026 & article & Generative AI & 40 & 44.5 & engineering & off\_scope \\
100+4406199489 & 2025 & article & Generative AI & 278 & 241.4 & survey & ok \\
101+4398243313 & 2024 & article & Generative AI & 463 & 139.4 & survey & ok \\
102+4409269614 & 2025 & review & Generative AI & 271 & 14.2 & survey & ok \\
103+3193917007 & 2021 & book-chapter & Generative AI & 195 & 127.1 & survey & ok \\
104+4288099666 & 2022 & preprint & Generative AI & 575 & 124.8 & method & ok \\
105+4394934695 & 2024 & article & Generative AI & 394 & 118.6 & empirical & ok \\
106+4321648656 & 2026 & article & Generative AI & 33 & 88.5 & survey & ok \\
107+4395957952 & 2023 & book & Generative AI & 400 & 116.7 & survey & no\_abstract \\
108+4394863715 & 2024 & preprint & Generative AI & 131 & 114.3 & empirical & no\_abstract \\
109+4407778399 & 2025 & article & Generative AI & 213 & 184.9 & empirical & ok \\
110+7124432017 & 2026 & article & Generative AI & 30 & 213.1 & theory & ok \\
111+7140170686 & 2026 & article & Generative AI & 30 & 213.1 & empirical & ok \\
112+4406728221 & 2025 & article & AI agents & 694 & 236.4 & survey & ok \\
113+7161734976 & 2026 & article & AI agents & 62 & 175.1 & engineering & ok \\
114+7135409568 & 2026 & article & AI agents & 52 & 23.7 & survey & ok \\
115+7134187809 & 2026 & article & AI agents & 46 & 129.9 & theory & off\_scope \\
116+4390730779 & 2026 & article & AI agents & 44 & 49.0 & survey & ok \\
117+4225355126 & 2022 & book-chapter & AI agents & 268 & 143.6 & survey & no\_abstract \\
118+4407570917 & 2026 & article & AI agents & 40 & 44.5 & engineering & ok \\
119+4410197190 & 2025 & review & AI agents & 267 & 33.1 & survey & ok \\
120+4413427262 & 2025 & article & AI agents & 258 & 87.9 & survey & ok \\
121+4414991289 & 2026 & article & AI agents & 31 & 23.9 & engineering & ok \\
122+7122857583 & 2026 & article & AI agents & 30 & 33.4 & theory & ok \\
123+4401023556 & 2024 & conference-paper & AI agents & 371 & 103.1 & survey & ok \\
124+4414023998 & 2025 & article & AI agents & 189 & 64.4 & survey & ok \\
125+7118083939 & 2026 & article & AI agents & 26 & 32.7 & theory & ok \\
126+4412505619 & 2025 & article & AI agents & 181 & 61.7 & survey & ok \\
127+4416209823 & 2025 & article & AI agents & 165 & 56.2 & survey & ok \\
128+7155930820 & 2026 & article & AI agents & 23 & 80.7 & theory & off\_scope \\
129+7162257503 & 2026 & article & AI agents & 22 & 77.2 & theory & off\_scope \\
130+4412709685 & 2025 & article & AI agents & 150 & 33.5 & engineering & no\_abstract \\
131+7128165235 & 2026 & article & AI agents & 21 & 23.4 & theory & ok \\
132+7118190454 & 2026 & article & AI agents & 21 & 22.2 & survey & ok \\
133+7130481397 & 2026 & article & AI agents & 21 & 9.6 & benchmark & ok \\
134+4390638955 & 2024 & article & AI agents & 240 & 72.3 & survey & off\_scope \\
135+4412203333 & 2025 & article & AI agents & 139 & 47.4 & survey & ok \\
136+4399982941 & 2024 & dissertation & AI agents & 8 & 68.2 & theory & normalisation\_artefact \\
137+2911283634 & 2020 & book-chapter & Reinforcement learning & 532 & 276.0 & survey & no\_abstract \\
138+4389473504 & 2023 & dissertation & Reinforcement learning & 28 & 232.9 & method & ok \\
139+2964083594 & 2025 & conference-paper & Reinforcement learning & 414 & 367.9 & empirical & year\_implausible \\
140+2098774185 & 2018 & article & Reinforcement learning & 1{,}077 & 203.3 & method & year\_implausible \\
141+1591675293 & 2025 & book-chapter & Reinforcement learning & 395 & 665.0 & method & year\_implausible \\
142+7123365826 & 2026 & article & Reinforcement learning & 55 & 155.3 & theory & off\_scope \\
143+2991046523 & 2021 & book-chapter & Reinforcement learning & 530 & 178.4 & survey & no\_abstract \\
144+2973229164 & 2019 & dissertation & Reinforcement learning & 47 & 171.2 & method & ok \\
145+2982316857 & 2019 & article & Reinforcement learning & 1{,}021 & 155.2 & engineering & no\_abstract \\
146+2902907165 & 2018 & article & Reinforcement learning & 733 & 138.3 & method & ok \\
147+7133481537 & 2026 & article & Reinforcement learning & 35 & 122.9 & theory & off\_scope \\
148+7133226216 & 2022 & conference-paper & Reinforcement learning & 482 & 95.8 & empirical & ok \\
149+3041807988 & 2020 & article & Reinforcement learning & 322 & 93.8 & theory & off\_scope \\
150+2781726626 & 2018 & preprint & Reinforcement learning & 775 & 90.8 & method & ok \\
151+2963923407 & 2018 & article & Reinforcement learning & 472 & 89.1 & method & ok \\
152+7122543576 & 2026 & article & Reinforcement learning & 25 & 70.6 & theory & off\_scope \\
153+2899138465 & 2018 & dissertation & Reinforcement learning & 20 & 88.6 & method & ok \\
154+4391631327 & 2024 & preprint & Reinforcement learning & 99 & 86.4 & empirical & identifier\_mismatch \\
155+4409147637 & 2025 & article & Reinforcement learning & 170 & 57.9 & method & ok \\
156+2895974005 & 2018 & dissertation & Reinforcement learning & 18 & 79.8 & empirical & normalisation\_artefact \\
157+7134057781 & 2026 & article & Reinforcement learning & 21 & 23.4 & engineering & ok \\
158+2754517384 & 2018 & conference-paper & Reinforcement learning & 452 & 70.0 & empirical & ok \\
159+2761873684 & 2018 & conference-paper & Reinforcement learning & 442 & 68.4 & method & ok \\
160+7133208539 & 2023 & conference-paper & Reinforcement learning & 293 & 67.6 & method & no\_abstract \\
161+7128436456 & 2026 & article & Reinforcement learning & 19 & 21.0 & survey & ok \\
162+2962862931 & 2024 & conference-paper & ML theory & 12{,}586 & 3498.4 & method & identifier\_mismatch \\
163+1698155719 & 2020 & book-chapter & ML theory & 1{,}212 & 971.7 & survey & ok \\
164+4288283362 & 2026 & article & ML theory & 174 & 576.9 & method & normalisation\_artefact \\
165+46659105 & 2020 & preprint & ML theory & 3{,}906 & 608.2 & theory & abstract\_mismatch \\
166+3145506661 & 2019 & book-chapter & ML theory & 1{,}581 & 491.8 & survey & no\_abstract \\
167+2291445757 & 2024 & book-chapter & ML theory & 769 & 426.8 & survey & ok \\
168+2100415632 & 2026 & preprint & ML theory & 116 & 667.5 & theory & year\_implausible \\
169+1568555062 & 2022 & book-chapter & ML theory & 576 & 308.7 & theory & ok \\
170+2751318774 & 2019 & book-chapter & ML theory & 978 & 304.2 & survey & ok \\
171+3208688400 & 2021 & book-chapter & ML theory & 833 & 280.5 & survey & no\_abstract \\
172+3105488314 & 2020 & dissertation & ML theory & 59 & 272.4 & theory & no\_abstract \\
173+57706852 & 2026 & preprint & ML theory & 72 & 414.3 & theory & year\_implausible \\
174+2946444988 & 2026 & preprint & ML theory & 68 & 2646.0 & method & year\_implausible \\
175+2945976633 & 2019 & article & ML theory & 1{,}556 & 236.6 & theory & no\_abstract \\
176+4400823820 & 2024 & article & ML theory & 782 & 235.4 & survey & ok \\
177+2999309192 & 2020 & article & ML theory & 1{,}781 & 222.8 & theory & ok \\
178+2999615587 & 2020 & article & ML theory & 1{,}767 & 221.0 & method & no\_abstract \\
179+2439568532 & 2018 & article & ML theory & 1{,}085 & 204.8 & theory & ok \\
180+3194464139 & 2020 & dissertation & ML theory & 21 & 197.5 & engineering & normalisation\_artefact \\
181+3041133507 & 2020 & article & ML theory & 1{,}538 & 192.4 & survey & ok \\
182+2436219157 & 2018 & article & ML theory & 956 & 180.4 & survey & ok \\
183+3135028703 & 2021 & article & ML theory & 1{,}518 & 176.9 & survey & no\_abstract \\
184+7124426637 & 2026 & article & ML theory & 46 & 51.2 & survey & ok \\
185+2165856466 & 2026 & preprint & ML theory & 45 & 259.0 & theory & year\_implausible \\
186+4402843978 & 2020 & report & ML theory & 455 & 158.3 & survey & ok \\
187+2964121744 & 2026 & article & DL architectures & 45{,}775 & 50938.5 & engineering & identifier\_mismatch \\
188+1710476689 & 2022 & book-chapter & DL architectures & 2{,}385 & 1018.2 & survey & ok \\
189+4322614756 & 2025 & conference-paper & DL architectures & 1{,}542 & 1370.5 & method & identifier\_mismatch \\
190+2963399829 & 2024 & preprint & DL architectures & 773 & 674.3 & method & year\_implausible \\
191+4382203079 & 2023 & conference-paper & DL architectures & 2{,}278 & 621.0 & empirical & ok \\
192+4406170795 & 2025 & article & DL architectures & 974 & 331.8 & method & ok \\
193+146900863 & 2018 & book-chapter & DL architectures & 999 & 464.0 & survey & no\_abstract \\
194+4396821501 & 2024 & preprint & DL architectures & 531 & 463.2 & method & ok \\
195+2950067852 & 2026 & preprint & DL architectures & 119 & 4630.6 & method & year\_implausible \\
196+4221150249 & 2026 & article & DL architectures & 107 & 119.1 & survey & ok \\
197+7148890333 & 2026 & article & DL architectures & 102 & 113.0 & method & no\_abstract \\
198+3177318507 & 2021 & conference-paper & DL architectures & 1{,}344 & 315.3 & method & ok \\
199+4389326242 & 2023 & preprint & DL architectures & 912 & 303.2 & method & ok \\
200+7118302872 & 2026 & article & DL architectures & 85 & 87.6 & method & ok \\
201+3137147200 & 2022 & book-chapter & DL architectures & 698 & 298.0 & survey & ok \\
202+3140854437 & 2021 & article & DL architectures & 2{,}452 & 293.1 & survey & ok \\
203+4210257598 & 2020 & article & DL architectures & 2{,}239 & 280.0 & survey & ok \\
204+2907492528 & 2020 & article & DL architectures & 2{,}230 & 278.9 & survey & ok \\
205+3152893301 & 2020 & article & DL architectures & 2{,}023 & 253.0 & survey & ok \\
206+7110333790 & 2025 & article & DL architectures & 496 & 169.0 & method & identifier\_mismatch \\
207+4385763767 & 2023 & conference-paper & DL architectures & 921 & 251.1 & survey & ok \\
208+3124006607 & 2026 & conference-paper & DL architectures & 69 & 529.6 & method & year\_implausible \\
209+4285723986 & 2020 & book-chapter & DL architectures & 471 & 244.4 & survey & no\_abstract \\
210+2892035503 & 2019 & article & DL architectures & 908 & 192.7 & survey & no\_abstract \\
211+2962776659 & 2024 & preprint & DL architectures & 212 & 184.9 & empirical & year\_implausible \\
212+4406063062 & 2025 & article & AI safety and alignment & 534 & 74.5 & survey & ok \\
213+7130596601 & 2026 & article & AI safety and alignment & 65 & 29.6 & survey & ok \\
214+3084163594 & 2020 & article & AI safety and alignment & 775 & 225.7 & survey & off\_scope \\
215+2963808661 & 2025 & conference-paper & AI safety and alignment & 390 & 126.3 & engineering & identifier\_mismatch \\
216+7133236347 & 2026 & article & AI safety and alignment & 51 & 44.5 & survey & ok \\
217+7128782475 & 2026 & article & AI safety and alignment & 51 & 23.2 & empirical & ok \\
218+4406571986 & 2025 & article & AI safety and alignment & 347 & 301.3 & empirical & ok \\
219+2981731882 & 2019 & article & AI safety and alignment & 1{,}097 & 166.8 & survey & no\_abstract \\
220+4402698709 & 2026 & article & AI safety and alignment & 44 & 38.4 & survey & ok \\
221+4408314709 & 2025 & article & AI safety and alignment & 303 & 42.3 & survey & ok \\
222+4407243994 & 2025 & article & AI safety and alignment & 300 & 41.8 & theory & ok \\
223+4330336443 & 2023 & preprint & AI safety and alignment & 495 & 146.6 & empirical & abstract\_mismatch \\
224+7128600296 & 2026 & review & AI safety and alignment & 41 & 12.8 & survey & ok \\
225+4399450035 & 2024 & conference-paper & AI safety and alignment & 359 & 138.5 & empirical & ok \\
226+4318586239 & 2023 & book-chapter & AI safety and alignment & 247 & 136.6 & survey & ok \\
227+4360615722 & 2023 & article & AI safety and alignment & 557 & 130.5 & theory & ok \\
228+2670253439 & 2018 & article & AI safety and alignment & 676 & 127.6 & survey & no\_abstract \\
229+7118631510 & 2026 & article & AI safety and alignment & 35 & 38.9 & method & no\_abstract \\
230+7127143535 & 2026 & article & AI safety and alignment & 33 & 50.0 & empirical & no\_abstract \\
231+4409249263 & 2025 & article & AI safety and alignment & 218 & 14.5 & survey & ok \\
232+4366262984 & 2023 & article & AI safety and alignment & 1{,}197 & 109.9 & survey & ok \\
233+2969896603 & 2021 & article & AI safety and alignment & 1{,}794 & 102.3 & survey & ok \\
234+4406152291 & 2025 & article & AI safety and alignment & 199 & 67.8 & empirical & ok \\
235+4410609100 & 2025 & conference-paper & AI safety and alignment & 193 & 171.5 & method & ok \\
236+4411542009 & 2025 & conference-paper & AI safety and alignment & 191 & 61.8 & empirical & ok \\
237+4360620450 & 2023 & article & Large language models & 3{,}137 & 113.5 & survey & ok \\
238+3185341429 & 2022 & article & Large language models & 2{,}462 & 253.7 & survey & ok \\
239+4384071683 & 2023 & article & Large language models & 2{,}448 & 38.1 & benchmark & ok \\
240+4384918448 & 2023 & preprint & Large language models & 2{,}431 & 88.5 & empirical & identifier\_mismatch \\
241+4393065402 & 2024 & article & Large language models & 1{,}820 & 226.1 & survey & ok \\
242+4319083882 & 2023 & article & Large language models & 1{,}590 & 146.0 & theory & no\_abstract \\
243+4399790131 & 2024 & review & AI safety and alignment & 1{,}484 & 81.8 & survey & ok \\
244+3133702157 & 2021 & conference-paper & Large language models & 1{,}464 & 232.1 & theory & ok \\
245+4321499901 & 2023 & article & Large language models & 1{,}389 & 325.5 & empirical & ok \\
246+\bottomrule
247+\multicolumn{8}{p{0.95\linewidth}}{\footnotesize Selection score and windows as in Part IV (Section~\ref{sec:m-score}); c36 and NCS36 as stored in the landmark file.}\\
248+\end{longtable}
modified manuscript/tables/tab_breaks.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/breaks.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv
2 \begin{longtable}{lrrrrrrrr} 2 \begin{longtable}{lrrrrrrrr}
3 \caption{Structural-change estimates by series: ITS level and slope changes at 2022-12 (HAC p-values), placebo p-values, Bai–Perron (BIC) and PELT break dates.}\label{tab:breaks}\\ 3 \caption{Structural-change estimates by series: ITS level and slope changes at 2022-12 (HAC p-values), placebo p-values, Bai–Perron (BIC) and PELT break dates.}\label{tab:breaks}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_cited_age.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/cited_age_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/cited_age_yearly.csv
2 \begin{table}[htbp]\centering\caption{Age of cited works by citing year (resolved references of eligible works; corpus version of the result file at generation time).}\label{tab:cited_age} 2 \begin{table}[htbp]\centering\caption{Age of cited works by citing year (resolved references of eligible works; corpus version of the result file at generation time).}\label{tab:cited_age}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrr} 4 \begin{tabular}{lrrrrrrr}
modified manuscript/tables/tab_collaboration.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/team_size_yearly.csv, ./results/bibliometrics/international_collab_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/team_size_yearly.csv, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/international_collab_yearly.csv
2 \begin{table}[htbp]\centering\caption{Team size and collaboration indicators by publication year (all eligible works; periods of the source file summed).}\label{tab:collaboration} 2 \begin{table}[htbp]\centering\caption{Team size and collaboration indicators by publication year (all eligible works; periods of the source file summed).}\label{tab:collaboration}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrr} 4 \begin{tabular}{lrrrrrrrr}
modified manuscript/tables/tab_countries.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/countries_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/countries_yearly.csv
2 \begin{table}[htbp]\centering\caption{Fractional output of the fifteen largest countries by year.}\label{tab:countries} 2 \begin{table}[htbp]\centering\caption{Fractional output of the fifteen largest countries by year.}\label{tab:countries}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_country_concentration.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/countries_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/countries_yearly.csv
2 \begin{table}[htbp]\centering\caption{Concentration of output across countries by year, and the two largest producers under full counting.}\label{tab:country_concentration} 2 \begin{table}[htbp]\centering\caption{Concentration of output across countries by year, and the two largest producers under full counting.}\label{tab:country_concentration}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_country_network.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/country_pairs_by_period.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/country_pairs_by_period.csv
2 \begin{table}[htbp]\centering\caption{The country co-authorship network by period: edges with at least fifty joint works.}\label{tab:country_network} 2 \begin{table}[htbp]\centering\caption{The country co-authorship network by period: edges with at least fifty joint works.}\label{tab:country_network}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrr} 4 \begin{tabular}{lrrrrrr}
modified manuscript/tables/tab_country_pairs.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/country_pairs_by_period.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/country_pairs_by_period.csv
2 \begin{table}[htbp]\centering\caption{The twenty strongest country pairs by period (number of joint works).}\label{tab:country_pairs} 2 \begin{table}[htbp]\centering\caption{The twenty strongest country pairs by period (number of joint works).}\label{tab:country_pairs}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrr} 4 \begin{tabular}{lrrrr}
modified manuscript/tables/tab_country_shares.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/countries_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/countries_yearly.csv
2 \begin{table}[htbp]\centering\caption{Share of fractional output of the fifteen largest countries by year (per cent).}\label{tab:country_shares} 2 \begin{table}[htbp]\centering\caption{Share of fractional output of the fifteen largest countries by year (per cent).}\label{tab:country_shares}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_dblp_only.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/dblp_only_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/dblp_only_yearly.csv
2 \begin{table}[htbp]\centering\caption{DBLP records of 2018--2026 without an OpenAlex counterpart, by DBLP record type and year.}\label{tab:dblp_only} 2 \begin{table}[htbp]\centering\caption{DBLP records of 2018--2026 without an OpenAlex counterpart, by DBLP record type and year.}\label{tab:dblp_only}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrr} 4 \begin{tabular}{lrrrrrrrr}
modified manuscript/tables/tab_emerging_h12.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/emerging/metrics_h12.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h12.json
2 \begin{table}[htbp]\centering\caption{Out-of-time performance of the emerging-topic detectors at horizon 12 months (test cutoffs 2023-01–2024-10).}\label{tab:emerging_h12} 2 \begin{table}[htbp]\centering\caption{Out-of-time performance of the emerging-topic detectors at horizon 12 months (test cutoffs 2023-01–2024-10).}\label{tab:emerging_h12}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrr} 4 \begin{tabular}{lrrr}
modified manuscript/tables/tab_emerging_h24.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/emerging/metrics_h24.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h24.json
2 \begin{table}[htbp]\centering\caption{Out-of-time performance of the emerging-topic detectors at horizon 24 months (test cutoffs 2023-01–2024-04).}\label{tab:emerging_h24} 2 \begin{table}[htbp]\centering\caption{Out-of-time performance of the emerging-topic detectors at horizon 24 months (test cutoffs 2023-01–2024-04).}\label{tab:emerging_h24}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrr} 4 \begin{tabular}{lrrr}
modified manuscript/tables/tab_field_flows.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/field_flows_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/field_flows_yearly.csv
2 \begin{table}[htbp]\centering\caption{Destination of the resolved references of eligible works by citing year: inside the corpus, computer science outside it, and the seven largest other OpenAlex fields (per cent of resolved references).}\label{tab:field_flows} 2 \begin{table}[htbp]\centering\caption{Destination of the resolved references of eligible works by citing year: inside the corpus, computer science outside it, and the seven largest other OpenAlex fields (per cent of resolved references).}\label{tab:field_flows}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrrr}
modified manuscript/tables/tab_indexing_lag.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/indexing_lag.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/indexing_lag.csv
2 \begin{table}[htbp]\centering\caption{The OpenAlex \texttt{created\_date} relative to the publication date, by publication year.}\label{tab:indexing_lag} 2 \begin{table}[htbp]\centering\caption{The OpenAlex \texttt{created\_date} relative to the publication date, by publication year.}\label{tab:indexing_lag}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrr} 4 \begin{tabular}{lrrrrr}
modified manuscript/tables/tab_institution_concentration.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/top_institutions_by_period.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_institutions_by_period.csv
2 \begin{table}[htbp]\centering\caption{Institutions with at least two hundred works per period: coverage and concentration (fractional counting).}\label{tab:institution_concentration} 2 \begin{table}[htbp]\centering\caption{Institutions with at least two hundred works per period: coverage and concentration (fractional counting).}\label{tab:institution_concentration}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrr} 4 \begin{tabular}{lrrrrrr}
modified manuscript/tables/tab_institution_types.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/institution_types_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/institution_types_yearly.csv
2 \begin{table}[htbp]\centering\caption{Share of typed affiliations by OpenAlex institution type and year (per cent of affiliation-work pairs).}\label{tab:institution_types} 2 \begin{table}[htbp]\centering\caption{Share of typed affiliations by OpenAlex institution type and year (per cent of affiliation-work pairs).}\label{tab:institution_types}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_known_set.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/emerging/known_set_lead_times.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/known_set_lead_times.csv
2 \begin{table}[htbp]\centering\caption{Preregistered known-emergence set: detection month, peak-growth month and lead time.}\label{tab:known_set} 2 \begin{table}[htbp]\centering\caption{Preregistered known-emergence set: detection month, peak-growth month and lead time.}\label{tab:known_set}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrr} 4 \begin{tabular}{lrrrr}
modified manuscript/tables/tab_p9_ai_shares.tex +2 −2
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/category_share_ci_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/category_share_ci_yearly.csv
2 \begin{table}[htbp]\centering\caption{Share of primary-category assignments held by the AI and machine-learning categories in 2018, 2022 and 2025 (per cent, bootstrap 95\% intervals).}\label{tab:p9-ai-shares} 2 \begin{table}[htbp]\centering\caption{Share of primary-category assignments held by the AI and machine-learning categories in 2018, 2022 and 2025 (per cent, bootstrap 95\% intervals).}\label{tab:p9-ai-shares}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrr} 4 \begin{tabular}{lrrr}
@@ -14,7 +14,7 @@ Reinforcement Learning & 0.66 [0.64, 0.68] & 1.25 [1.22, 1.28] & 1.48 [1.46, 1.5
14 Machine Learning Theory & 2.71 [2.67, 2.76] & 3.59 [3.54, 3.64] & 3.67 [3.63, 3.71] \\ 14 Machine Learning Theory & 2.71 [2.67, 2.76] & 3.59 [3.54, 3.64] & 3.67 [3.63, 3.71] \\
15 Deep Learning Architectures & 2.40 [2.36, 2.44] & 3.97 [3.93, 4.02] & 3.82 [3.79, 3.86] \\ 15 Deep Learning Architectures & 2.40 [2.36, 2.44] & 3.97 [3.93, 4.02] & 3.82 [3.79, 3.86] \\
16 AI Safety and Alignment & 0.21 [0.20, 0.22] & 0.53 [0.51, 0.55] & 1.22 [1.19, 1.24] \\ 16 AI Safety and Alignment & 0.21 [0.20, 0.22] & 0.53 [0.51, 0.55] & 1.22 [1.19, 1.24] \\
17−AI \\& ML area (sum) & 7.65 & 11.41 & 18.63 \\ 17+AI \& ML area (sum) & 7.65 & 11.41 & 18.63 \\
18 \bottomrule 18 \bottomrule
19 \end{tabular} 19 \end{tabular}
20 \begin{tablenotes}\footnotesize\item Shares are of the year's works with a primary category; intervals from 500 multinomial resamples of the works (seed 44). The area row is the sum of the category shares (no interval). Categories 1--8 and 38 of the mandate.\end{tablenotes} 20 \begin{tablenotes}\footnotesize\item Shares are of the year's works with a primary category; intervals from 500 multinomial resamples of the works (seed 44). The area row is the sum of the category shares (no interval). Categories 1--8 and 38 of the mandate.\end{tablenotes}
modified manuscript/tables/tab_p9_altdates.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{longtable}{lrrrrr} 2 \begin{longtable}{lrrrrr}
3 \caption{ITS level change (HAC $p$) by category at the five pre-specified event dates.}\label{tab:p9-altdates}\\ 3 \caption{ITS level change (HAC $p$) by category at the five pre-specified event dates.}\label{tab:p9-altdates}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_p9_area_flow_change.tex +8 −8
@@ -1,17 +1,17 @@
1−% generated by tables/make_tables.py from ./results/graph/area_flows_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/area_flows_yearly.csv
2 \begin{table}[htbp]\centering\caption{Change in the area citation-flow matrix between 2018--2022 and 2023--2026: post-period row share minus baseline row share (percentage points).}\label{tab:p9-area-flow-change} 2 \begin{table}[htbp]\centering\caption{Change in the area citation-flow matrix between 2018--2022 and 2023--2026: post-period row share minus baseline row share (percentage points).}\label{tab:p9-area-flow-change}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrr} 4 \begin{tabular}{lrrrrrrrr}
5 \toprule 5 \toprule
6−\textbf{Citing area} & \textbf{AI \\& ML} & \textbf{Perception \\& interaction} & \textbf{Software \\& PL} & \textbf{Theory} & \textbf{Data \\& distributed} & \textbf{Security \\& privacy} & \textbf{Hardware \\& emerging} & \textbf{Computational science} \\ 6+\textbf{Citing area} & \textbf{AI \& ML} & \textbf{Perception \& interaction} & \textbf{Software \& PL} & \textbf{Theory} & \textbf{Data \& distributed} & \textbf{Security \& privacy} & \textbf{Hardware \& emerging} & \textbf{Computational science} \\
7 \midrule 7 \midrule
8−AI \\& ML & -3.1 & +3.1 & +0.2 & -0.9 & +0.1 & +0.5 & -0.2 & +0.2 \\ 8+AI \& ML & -3.1 & +3.1 & +0.2 & -0.9 & +0.1 & +0.5 & -0.2 & +0.2 \\
9−Perception \\& interaction & +0.4 & +0.2 & -0.1 & -0.0 & -0.4 & +0.0 & -0.0 & +0.0 \\ 9+Perception \& interaction & +0.4 & +0.2 & -0.1 & -0.0 & -0.4 & +0.0 & -0.0 & +0.0 \\
10−Software \\& PL & +2.6 & +1.2 & -2.8 & -0.6 & -1.0 & +0.7 & -0.0 & -0.2 \\ 10+Software \& PL & +2.6 & +1.2 & -2.8 & -0.6 & -1.0 & +0.7 & -0.0 & -0.2 \\
11 Theory & +1.1 & +1.1 & -0.5 & -1.0 & -0.6 & -0.0 & +0.2 & -0.2 \\ 11 Theory & +1.1 & +1.1 & -0.5 & -1.0 & -0.6 & -0.0 & +0.2 & -0.2 \\
12−Data \\& distributed & +1.5 & +0.7 & -0.0 & -0.1 & -3.0 & +1.0 & +0.1 & -0.1 \\ 12+Data \& distributed & +1.5 & +0.7 & -0.0 & -0.1 & -3.0 & +1.0 & +0.1 & -0.1 \\
13−Security \\& privacy & +1.8 & +0.9 & +0.0 & -0.1 & +0.1 & -2.6 & -0.1 & +0.0 \\ 13+Security \& privacy & +1.8 & +0.9 & +0.0 & -0.1 & +0.1 & -2.6 & -0.1 & +0.0 \\
14−Hardware \\& emerging & +0.4 & -0.2 & -0.2 & +0.0 & +0.1 & +0.1 & -0.2 & -0.0 \\ 14+Hardware \& emerging & +0.4 & -0.2 & -0.2 & +0.0 & +0.1 & +0.1 & -0.2 & -0.0 \\
15 Computational science & +2.5 & +0.9 & -0.1 & -0.7 & -0.7 & +0.0 & -0.1 & -1.8 \\ 15 Computational science & +2.5 & +0.9 & -0.1 & -0.7 & -0.7 & +0.0 & -0.1 & -1.8 \\
16 \bottomrule 16 \bottomrule
17 \end{tabular} 17 \end{tabular}
modified manuscript/tables/tab_p9_area_shares.tex +2 −2
@@ -1,9 +1,9 @@
1−% generated by tables/make_tables.py from ./results/graph/category_counts_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv
2 \begin{table}[htbp]\centering\caption{Share of primary-category assignments by top-level area and year (per cent), with the two protocol periods.}\label{tab:p9-area-shares} 2 \begin{table}[htbp]\centering\caption{Share of primary-category assignments by top-level area and year (per cent), with the two protocol periods.}\label{tab:p9-area-shares}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrr} 4 \begin{tabular}{lrrrrrrrr}
5 \toprule 5 \toprule
6−\textbf{Year / period} & \textbf{AI \\& ML} & \textbf{Perception \\& interaction} & \textbf{Software \\& PL} & \textbf{Theory} & \textbf{Data \\& distributed} & \textbf{Security \\& privacy} & \textbf{Hardware \\& emerging} & \textbf{Computational science} \\ 6+\textbf{Year / period} & \textbf{AI \& ML} & \textbf{Perception \& interaction} & \textbf{Software \& PL} & \textbf{Theory} & \textbf{Data \& distributed} & \textbf{Security \& privacy} & \textbf{Hardware \& emerging} & \textbf{Computational science} \\
7 \midrule 7 \midrule
8 2018 & 7.65 & 33.45 & 7.89 & 6.60 & 22.66 & 9.99 & 5.75 & 5.99 \\ 8 2018 & 7.65 & 33.45 & 7.89 & 6.60 & 22.66 & 9.99 & 5.75 & 5.99 \\
9 2019 & 8.89 & 34.69 & 7.21 & 6.15 & 21.14 & 10.27 & 5.48 & 6.17 \\ 9 2019 & 8.89 & 34.69 & 7.21 & 6.15 & 21.14 & 10.27 & 5.48 & 6.17 \\
modified manuscript/tables/tab_p9_category_periods.tex +36 −36
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_counts_yearly.csv, ./results/bibliometrics/cagr_by_category.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/cagr_by_category.csv
2 \begin{longtable}{lrrrrrrrr} 2 \begin{longtable}{lrrrrrrrr}
3 \caption{The forty categories by protocol period: primary-category output, composition shares, multi-label growth and compound annual growth rates.}\label{tab:p9-category-periods}\\ 3 \caption{The forty categories by protocol period: primary-category output, composition shares, multi-label growth and compound annual growth rates.}\label{tab:p9-category-periods}\\
4 \toprule 4 \toprule
@@ -9,45 +9,45 @@
9 \textbf{Category} & \textbf{Area} & \textbf{Primary, base} & \textbf{Primary, post} & \textbf{Share base (\%)} & \textbf{Share post (\%)} & \textbf{Multi-label post/base} & \textbf{CAGR base} & \textbf{CAGR post} \\ 9 \textbf{Category} & \textbf{Area} & \textbf{Primary, base} & \textbf{Primary, post} & \textbf{Share base (\%)} & \textbf{Share post (\%)} & \textbf{Multi-label post/base} & \textbf{CAGR base} & \textbf{CAGR post} \\
10 \midrule 10 \midrule
11 \endhead 11 \endhead
12−Computer Vision & Perception \\& interaction & 421{,}385 & 521{,}935 & 15.79 & 16.97 & 1.30 & 0.091 & 0.149 \\ 12+Computer Vision & Perception \& interaction & 421{,}385 & 521{,}935 & 15.79 & 16.97 & 1.30 & 0.091 & 0.149 \\
13−Computer Networks & Data \\& distributed & 267{,}896 & 212{,}342 & 10.04 & 6.90 & 0.94 & -0.044 & 0.071 \\ 13+Computer Networks & Data \& distributed & 267{,}896 & 212{,}342 & 10.04 & 6.90 & 0.94 & -0.044 & 0.071 \\
14−Human-Computer Interaction & Perception \\& interaction & 174{,}138 & 196{,}794 & 6.53 & 6.40 & 1.26 & 0.004 & 0.197 \\ 14+Human-Computer Interaction & Perception \& interaction & 174{,}138 & 196{,}794 & 6.53 & 6.40 & 1.26 & 0.004 & 0.197 \\
15−Cybersecurity & Security \\& privacy & 154{,}197 & 195{,}498 & 5.78 & 6.36 & 1.26 & 0.017 & 0.272 \\ 15+Cybersecurity & Security \& privacy & 154{,}197 & 195{,}498 & 5.78 & 6.36 & 1.26 & 0.017 & 0.272 \\
16−Natural Language Processing & Perception \\& interaction & 134{,}844 & 152{,}754 & 5.05 & 4.97 & 1.75 & 0.062 & 0.124 \\ 16+Natural Language Processing & Perception \& interaction & 134{,}844 & 152{,}754 & 5.05 & 4.97 & 1.75 & 0.062 & 0.124 \\
17−Large Language Models & AI \\& ML & 11{,}813 & 132{,}127 & 0.44 & 4.30 & 3.49 & 0.189 & 1.085 \\ 17+Large Language Models & AI \& ML & 11{,}813 & 132{,}127 & 0.44 & 4.30 & 3.49 & 0.189 & 1.085 \\
18−Software Engineering & Software \\& PL & 132{,}205 & 129{,}300 & 4.95 & 4.20 & 1.32 & -0.049 & 0.177 \\ 18+Software Engineering & Software \& PL & 132{,}205 & 129{,}300 & 4.95 & 4.20 & 1.32 & -0.049 & 0.177 \\
19 Algorithms and Data Structures & Theory & 144{,}105 & 121{,}771 & 5.40 & 3.96 & 0.99 & -0.008 & 0.088 \\ 19 Algorithms and Data Structures & Theory & 144{,}105 & 121{,}771 & 5.40 & 3.96 & 0.99 & -0.008 & 0.088 \\
20−Robotics & Perception \\& interaction & 101{,}643 & 118{,}241 & 3.81 & 3.84 & 1.28 & 0.072 & 0.140 \\ 20+Robotics & Perception \& interaction & 101{,}643 & 118{,}241 & 3.81 & 3.84 & 1.28 & 0.072 & 0.140 \\
21−Deep Learning Architectures & AI \\& ML & 87{,}481 & 116{,}113 & 3.28 & 3.77 & 1.57 & 0.167 & 0.155 \\ 21+Deep Learning Architectures & AI \& ML & 87{,}481 & 116{,}113 & 3.28 & 3.77 & 1.57 & 0.167 & 0.155 \\
22−Machine Learning Theory & AI \\& ML & 88{,}923 & 112{,}830 & 3.33 & 3.67 & 1.51 & 0.104 & 0.197 \\ 22+Machine Learning Theory & AI \& ML & 88{,}923 & 112{,}830 & 3.33 & 3.67 & 1.51 & 0.104 & 0.197 \\
23 Computational Biology & Computational science & 82{,}846 & 97{,}856 & 3.10 & 3.18 & 1.31 & 0.108 & 0.170 \\ 23 Computational Biology & Computational science & 82{,}846 & 97{,}856 & 3.10 & 3.18 & 1.31 & 0.108 & 0.170 \\
24 Scientific Computing & Computational science & 93{,}944 & 84{,}877 & 3.52 & 2.76 & 1.10 & 0.042 & 0.066 \\ 24 Scientific Computing & Computational science & 93{,}944 & 84{,}877 & 3.52 & 2.76 & 1.10 & 0.042 & 0.066 \\
25−Computer Architecture & Hardware \\& emerging & 86{,}471 & 70{,}512 & 3.24 & 2.29 & 0.98 & -0.061 & 0.111 \\ 25+Computer Architecture & Hardware \& emerging & 86{,}471 & 70{,}512 & 3.24 & 2.29 & 0.98 & -0.061 & 0.111 \\
26−Speech and Audio Computing & Perception \\& interaction & 62{,}133 & 65{,}819 & 2.33 & 2.14 & 1.14 & 0.012 & 0.118 \\ 26+Speech and Audio Computing & Perception \& interaction & 62{,}133 & 65{,}819 & 2.33 & 2.14 & 1.14 & 0.012 & 0.118 \\
27−Blockchain Systems & Security \\& privacy & 46{,}516 & 62{,}087 & 1.74 & 2.02 & 1.40 & 0.163 & 0.120 \\ 27+Blockchain Systems & Security \& privacy & 46{,}516 & 62{,}087 & 1.74 & 2.02 & 1.40 & 0.163 & 0.120 \\
28−Information Retrieval & Data \\& distributed & 58{,}198 & 61{,}738 & 2.18 & 2.01 & 1.34 & -0.018 & 0.176 \\ 28+Information Retrieval & Data \& distributed & 58{,}198 & 61{,}738 & 2.18 & 2.01 & 1.34 & -0.018 & 0.176 \\
29−Database Systems & Data \\& distributed & 74{,}552 & 61{,}582 & 2.79 & 2.00 & 1.02 & -0.043 & 0.073 \\ 29+Database Systems & Data \& distributed & 74{,}552 & 61{,}582 & 2.79 & 2.00 & 1.02 & -0.043 & 0.073 \\
30−Quantum Computing & Hardware \\& emerging & 39{,}846 & 60{,}417 & 1.49 & 1.96 & 1.59 & 0.103 & 0.287 \\ 30+Quantum Computing & Hardware \& emerging & 39{,}846 & 60{,}417 & 1.49 & 1.96 & 1.59 & 0.103 & 0.287 \\
31−Applied Cryptography & Security \\& privacy & 53{,}877 & 51{,}175 & 2.02 & 1.66 & 1.13 & -0.011 & 0.141 \\ 31+Applied Cryptography & Security \& privacy & 53{,}877 & 51{,}175 & 2.02 & 1.66 & 1.13 & -0.011 & 0.141 \\
32−Reinforcement Learning & AI \\& ML & 26{,}913 & 44{,}477 & 1.01 & 1.45 & 2.08 & 0.208 & 0.252 \\ 32+Reinforcement Learning & AI \& ML & 26{,}913 & 44{,}477 & 1.01 & 1.45 & 2.08 & 0.208 & 0.252 \\
33−Distributed Computing & Data \\& distributed & 52{,}661 & 43{,}642 & 1.97 & 1.42 & 1.22 & -0.050 & 0.132 \\ 33+Distributed Computing & Data \& distributed & 52{,}661 & 43{,}642 & 1.97 & 1.42 & 1.22 & -0.050 & 0.132 \\
34−Computer Graphics & Perception \\& interaction & 46{,}021 & 43{,}270 & 1.72 & 1.41 & 1.20 & -0.051 & 0.157 \\ 34+Computer Graphics & Perception \& interaction & 46{,}021 & 43{,}270 & 1.72 & 1.41 & 1.20 & -0.051 & 0.157 \\
35−Edge Computing & Data \\& distributed & 34{,}568 & 40{,}445 & 1.30 & 1.31 & 1.28 & 0.133 & 0.126 \\ 35+Edge Computing & Data \& distributed & 34{,}568 & 40{,}445 & 1.30 & 1.31 & 1.28 & 0.133 & 0.126 \\
36−Generative AI & AI \\& ML & 7{,}503 & 38{,}194 & 0.28 & 1.24 & 3.44 & 0.296 & 0.870 \\ 36+Generative AI & AI \& ML & 7{,}503 & 38{,}194 & 0.28 & 1.24 & 3.44 & 0.296 & 0.870 \\
37−Cloud Computing & Data \\& distributed & 40{,}415 & 36{,}757 & 1.51 & 1.19 & 1.02 & -0.068 & 0.198 \\ 37+Cloud Computing & Data \& distributed & 40{,}415 & 36{,}757 & 1.51 & 1.19 & 1.02 & -0.068 & 0.198 \\
38−AI Safety and Alignment & AI \\& ML & 9{,}997 & 33{,}792 & 0.37 & 1.10 & 3.71 & 0.294 & 0.647 \\ 38+AI Safety and Alignment & AI \& ML & 9{,}997 & 33{,}792 & 0.37 & 1.10 & 3.71 & 0.294 & 0.647 \\
39−AI Agents and Autonomous Systems & AI \\& ML & 25{,}167 & 32{,}219 & 0.94 & 1.05 & 2.00 & -0.045 & 0.329 \\ 39+AI Agents and Autonomous Systems & AI \& ML & 25{,}167 & 32{,}219 & 0.94 & 1.05 & 2.00 & -0.045 & 0.329 \\
40−Privacy-Preserving Computation & Security \\& privacy & 18{,}816 & 28{,}450 & 0.71 & 0.92 & 1.99 & 0.090 & 0.327 \\ 40+Privacy-Preserving Computation & Security \& privacy & 18{,}816 & 28{,}450 & 0.71 & 0.92 & 1.99 & 0.090 & 0.327 \\
41−Multimodal AI & AI \\& ML & 4{,}007 & 20{,}970 & 0.15 & 0.68 & 3.75 & 0.284 & 0.869 \\ 41+Multimodal AI & AI \& ML & 4{,}007 & 20{,}970 & 0.15 & 0.68 & 3.75 & 0.284 & 0.869 \\
42−Formal Verification & Software \\& PL & 22{,}992 & 18{,}225 & 0.86 & 0.59 & 0.91 & -0.069 & 0.027 \\ 42+Formal Verification & Software \& PL & 22{,}992 & 18{,}225 & 0.86 & 0.59 & 0.91 & -0.069 & 0.027 \\
43−Programming Languages & Software \\& PL & 20{,}392 & 15{,}837 & 0.76 & 0.51 & 0.98 & -0.044 & 0.040 \\ 43+Programming Languages & Software \& PL & 20{,}392 & 15{,}837 & 0.76 & 0.51 & 0.98 & -0.044 & 0.040 \\
44 Computational Complexity & Theory & 18{,}778 & 15{,}760 & 0.70 & 0.51 & 1.09 & -0.020 & 0.097 \\ 44 Computational Complexity & Theory & 18{,}778 & 15{,}760 & 0.70 & 0.51 & 1.09 & -0.020 & 0.097 \\
45−Federated Learning & Security \\& privacy & 2{,}229 & 9{,}011 & 0.08 & 0.29 & 3.69 & 0.615 & 0.304 \\ 45+Federated Learning & Security \& privacy & 2{,}229 & 9{,}011 & 0.08 & 0.29 & 3.69 & 0.615 & 0.304 \\
46−Accelerators and AI Hardware & Hardware \\& emerging & 5{,}520 & 6{,}917 & 0.21 & 0.22 & 1.51 & 0.148 & 0.092 \\ 46+Accelerators and AI Hardware & Hardware \& emerging & 5{,}520 & 6{,}917 & 0.21 & 0.22 & 1.51 & 0.148 & 0.092 \\
47−Foundation Models & AI \\& ML & 838 & 6{,}186 & 0.03 & 0.20 & 3.12 & 0.661 & 0.802 \\ 47+Foundation Models & AI \& ML & 838 & 6{,}186 & 0.03 & 0.20 & 3.12 & 0.661 & 0.802 \\
48−Sustainable Computing & Hardware \\& emerging & 3{,}356 & 5{,}389 & 0.13 & 0.18 & 1.37 & 0.092 & 0.233 \\ 48+Sustainable Computing & Hardware \& emerging & 3{,}356 & 5{,}389 & 0.13 & 0.18 & 1.37 & 0.092 & 0.233 \\
49−Operating Systems & Hardware \\& emerging & 6{,}877 & 5{,}158 & 0.26 & 0.17 & 0.96 & -0.087 & 0.078 \\ 49+Operating Systems & Hardware \& emerging & 6{,}877 & 5{,}158 & 0.26 & 0.17 & 0.96 & -0.087 & 0.078 \\
50−Compilers & Software \\& PL & 3{,}450 & 2{,}900 & 0.13 & 0.09 & 1.20 & -0.030 & 0.053 \\ 50+Compilers & Software \& PL & 3{,}450 & 2{,}900 & 0.13 & 0.09 & 1.20 & -0.030 & 0.053 \\
51 Autonomous Scientific Discovery & Computational science & 841 & 2{,}539 & 0.03 & 0.08 & 2.50 & 0.287 & 0.382 \\ 51 Autonomous Scientific Discovery & Computational science & 841 & 2{,}539 & 0.03 & 0.08 & 2.50 & 0.287 & 0.382 \\
52 \bottomrule 52 \bottomrule
53 \multicolumn{9}{p{0.95\linewidth}}{\footnotesize Primary counts and shares use the primary category; the multi-label ratio is the post-period count of all assignments to the category divided by the baseline count (periods split at the reference date). CAGR base: 2018 to 2022 (four years); CAGR post: 2023 to 2025 (two years), both on primary counts of complete calendar years (equation~\eqref{eq:cagr}). Sorted by post-period share.}\\ 53 \multicolumn{9}{p{0.95\linewidth}}{\footnotesize Primary counts and shares use the primary category; the multi-label ratio is the post-period count of all assignments to the category divided by the baseline count (periods split at the reference date). CAGR base: 2018 to 2022 (four years); CAGR post: 2023 to 2025 (two years), both on primary counts of complete calendar years (equation~\eqref{eq:cagr}). Sorted by post-period share.}\\
modified manuscript/tables/tab_p9_classification.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_counts_yearly.csv, ./data/metadata/prisma_counts.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/data/metadata/prisma_counts.json
2 \begin{table}[htbp]\centering\caption{Classification coverage by publication year: eligible works, works with a primary category, and multi-label assignments.}\label{tab:p9-classification} 2 \begin{table}[htbp]\centering\caption{Classification coverage by publication year: eligible works, works with a primary category, and multi-label assignments.}\label{tab:p9-classification}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrr} 4 \begin{tabular}{lrrrrr}
modified manuscript/tables/tab_p9_cooccurrence.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_cooccurrence_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_cooccurrence_yearly.csv
2 \begin{table}[htbp]\centering\caption{The fifteen most frequent co-classification pairs (works carrying both categories in their label set), 2018--2022 and 2023--2026.}\label{tab:p9-cooccurrence} 2 \begin{table}[htbp]\centering\caption{The fifteen most frequent co-classification pairs (works carrying both categories in their label set), 2018--2022 and 2023--2026.}\label{tab:p9-cooccurrence}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrr} 4 \begin{tabular}{lrrrr}
modified manuscript/tables/tab_p9_cross_area.tex +7 −7
@@ -1,17 +1,17 @@
1−% generated by tables/make_tables.py from ./results/graph/cross_area_share_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/cross_area_share_yearly.csv
2 \begin{table}[htbp]\centering\caption{Cross-area share of internal references by citing area and citing year (per cent), with the aggregate share and its composition-adjusted counterpart.}\label{tab:p9-cross-area} 2 \begin{table}[htbp]\centering\caption{Cross-area share of internal references by citing area and citing year (per cent), with the aggregate share and its composition-adjusted counterpart.}\label{tab:p9-cross-area}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
5 \toprule 5 \toprule
6 \textbf{Citing area} & \textbf{2018} & \textbf{2019} & \textbf{2020} & \textbf{2021} & \textbf{2022} & \textbf{2023} & \textbf{2024} & \textbf{2025} & \textbf{2026} \\ 6 \textbf{Citing area} & \textbf{2018} & \textbf{2019} & \textbf{2020} & \textbf{2021} & \textbf{2022} & \textbf{2023} & \textbf{2024} & \textbf{2025} & \textbf{2026} \\
7 \midrule 7 \midrule
8−AI \\& ML & 46.6 & 48.1 & 49.3 & 51.0 & 55.1 & 55.3 & 54.9 & 55.4 & 53.9 \\ 8+AI \& ML & 46.6 & 48.1 & 49.3 & 51.0 & 55.1 & 55.3 & 54.9 & 55.4 & 53.9 \\
9 Computational science & 31.9 & 33.4 & 31.6 & 31.4 & 32.3 & 33.8 & 34.8 & 34.6 & 31.3 \\ 9 Computational science & 31.9 & 33.4 & 31.6 & 31.4 & 32.3 & 33.8 & 34.8 & 34.6 & 31.3 \\
10−Data \\& distributed & 25.8 & 24.2 & 25.1 & 26.7 & 27.1 & 28.1 & 28.9 & 29.7 & 30.8 \\ 10+Data \& distributed & 25.8 & 24.2 & 25.1 & 26.7 & 27.1 & 28.1 & 28.9 & 29.7 & 30.8 \\
11−Hardware \\& emerging & 30.2 & 29.0 & 29.2 & 28.5 & 28.0 & 28.8 & 28.7 & 28.1 & 29.6 \\ 11+Hardware \& emerging & 30.2 & 29.0 & 29.2 & 28.5 & 28.0 & 28.8 & 28.7 & 28.1 & 29.6 \\
12−Perception \\& interaction & 20.5 & 16.2 & 15.7 & 15.8 & 15.2 & 15.3 & 15.2 & 15.6 & 16.3 \\ 12+Perception \& interaction & 20.5 & 16.2 & 15.7 & 15.8 & 15.2 & 15.3 & 15.2 & 15.6 & 16.3 \\
13−Security \\& privacy & 32.6 & 34.4 & 35.8 & 36.7 & 36.7 & 37.3 & 38.8 & 39.6 & 39.9 \\ 13+Security \& privacy & 32.6 & 34.4 & 35.8 & 36.7 & 36.7 & 37.3 & 38.8 & 39.6 & 39.9 \\
14−Software \\& PL & 29.9 & 30.2 & 32.8 & 33.5 & 33.8 & 35.3 & 35.8 & 36.3 & 35.5 \\ 14+Software \& PL & 29.9 & 30.2 & 32.8 & 33.5 & 33.8 & 35.3 & 35.8 & 36.3 & 35.5 \\
15 Theory & 30.0 & 35.1 & 36.1 & 35.9 & 34.7 & 35.4 & 35.8 & 36.5 & 37.8 \\ 15 Theory & 30.0 & 35.1 & 36.1 & 35.9 & 34.7 & 35.4 & 35.8 & 36.5 & 37.8 \\
16 All areas, raw & 27.8 & 25.9 & 26.3 & 26.9 & 26.8 & 27.4 & 28.2 & 29.5 & 30.5 \\ 16 All areas, raw & 27.8 & 25.9 & 26.3 & 26.9 & 26.8 & 27.4 & 28.2 & 29.5 & 30.5 \\
17 All areas, baseline area weights & 27.4 & 25.9 & 26.1 & 26.8 & 27.1 & 27.6 & 27.8 & 28.3 & 28.6 \\ 17 All areas, baseline area weights & 27.4 & 25.9 & 26.1 & 26.8 & 27.1 & 27.6 & 27.8 & 28.3 & 28.6 \\
modified manuscript/tables/tab_p9_emerging_cutoffs.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/emerging/metrics_h12.json, ./results/emerging/metrics_h24.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h12.json, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h24.json
2 \begin{table}[htbp]\centering\caption{AUC of the three detectors at each test cutoff, by horizon.}\label{tab:p9-emerging-cutoffs} 2 \begin{table}[htbp]\centering\caption{AUC of the three detectors at each test cutoff, by horizon.}\label{tab:p9-emerging-cutoffs}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrr} 4 \begin{tabular}{lrrrrrr}
modified manuscript/tables/tab_p9_emerging_setup.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/emerging/metrics_h12.json, ./results/emerging/metrics_h24.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h12.json, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/emerging/metrics_h24.json
2 \begin{table}[htbp]\centering\caption{Out-of-time evaluation design of the emerging-topic detector: cutoffs, panel sizes and positive rates.}\label{tab:p9-emerging-setup} 2 \begin{table}[htbp]\centering\caption{Out-of-time evaluation design of the emerging-topic detector: cutoffs, panel sizes and positive rates.}\label{tab:p9-emerging-setup}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrr} 4 \begin{tabular}{lrrrrrr}
modified manuscript/tables/tab_p9_h3.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{table}[htbp]\centering\caption{Hypothesis H3: break dates and ITS estimates for the eight categories named in the protocol.}\label{tab:p9-h3} 2 \begin{table}[htbp]\centering\caption{Hypothesis H3: break dates and ITS estimates for the eight categories named in the protocol.}\label{tab:p9-h3}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrr} 4 \begin{tabular}{lrrrrrrrr}
modified manuscript/tables/tab_p9_its_levels.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{longtable}{lrrrrrrrr} 2 \begin{longtable}{lrrrrrrrr}
3 \caption{ITS level change at December 2022 by category: full-sample and donut estimates, HAC and placebo $p$-values, and the BOCPD change probability over December 2022 to March 2023.}\label{tab:p9-its-levels}\\ 3 \caption{ITS level change at December 2022 by category: full-sample and donut estimates, HAC and placebo $p$-values, and the BOCPD change probability over December 2022 to March 2023.}\label{tab:p9-its-levels}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_p9_its_slopes.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{longtable}{lrrrrrrrr} 2 \begin{longtable}{lrrrrrrrr}
3 \caption{ITS slope change at December 2022 by category: pre-period slope, change, HAC and placebo $p$-values, and implied post-period slope (monthly log growth).}\label{tab:p9-its-slopes}\\ 3 \caption{ITS slope change at December 2022 by category: pre-period slope, change, HAC and placebo $p$-values, and implied post-period slope (monthly log growth).}\label{tab:p9-its-slopes}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_p9_its_total_events.tex +2 −2
@@ -1,9 +1,9 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{table}[htbp]\centering\caption{Interrupted time series of the total monthly series at the reference date and at four alternative event dates (log counts, month-of-year dummies, Newey--West standard errors).}\label{tab:p9-its-total-events} 2 \begin{table}[htbp]\centering\caption{Interrupted time series of the total monthly series at the reference date and at four alternative event dates (log counts, month-of-year dummies, Newey--West standard errors).}\label{tab:p9-its-total-events}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrr} 4 \begin{tabular}{lrrrrrr}
5 \toprule 5 \toprule
6−\textbf{Event (month)} & \textbf{Pre-slope (SE)} & \textbf{Level change (SE)} & \textbf{HAC p} & \textbf{Slope change (SE)} & \textbf{HAC p } & \textbf{R2} \\ 6+\textbf{Event (month)} & \textbf{Pre-slope (SE)} & \textbf{Level change (SE)} & \textbf{HAC p (level)} & \textbf{Slope change (SE)} & \textbf{HAC p (slope)} & \textbf{$R^2$} \\
7 \midrule 7 \midrule
8 GPT-3 (2020-06) & 0.0034 (0.0012) & -0.091 (0.059) & 0.123 & +0.0059 (0.0015) & 8.51e-05 & 0.926 \\ 8 GPT-3 (2020-06) & 0.0034 (0.0012) & -0.091 (0.059) & 0.123 & +0.0059 (0.0015) & 8.51e-05 & 0.926 \\
9 DALL$\cdot$E 2 (2022-05) & 0.0034 (0.0007) & -0.075 (0.028) & 0.006 & +0.0100 (0.0009) & 1.39e-28 & 0.965 \\ 9 DALL$\cdot$E 2 (2022-05) & 0.0034 (0.0007) & -0.075 (0.028) & 0.006 & +0.0100 (0.0009) & 1.39e-28 & 0.965 \\
modified manuscript/tables/tab_p9_landmarks.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./literature_synthesis/landmarks/selection_report.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/landmarks/selection_report.json
2 \begin{table}[htbp]\centering\caption{Composition of the landmark set by area and by publication year.}\label{tab:p9-landmarks} 2 \begin{table}[htbp]\centering\caption{Composition of the landmark set by area and by publication year.}\label{tab:p9-landmarks}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrr} 4 \begin{tabular}{lrr}
modified manuscript/tables/tab_p9_llm_partners.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_cooccurrence_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_cooccurrence_yearly.csv
2 \begin{table}[htbp]\centering\caption{Categories most often co-assigned with Large Language Models, by span.}\label{tab:p9-llm-partners} 2 \begin{table}[htbp]\centering\caption{Categories most often co-assigned with Large Language Models, by span.}\label{tab:p9-llm-partners}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrr} 4 \begin{tabular}{lrrr}
modified manuscript/tables/tab_p9_modularity.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/modularity_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/modularity_yearly.csv
2 \begin{table}[htbp]\centering\caption{Modularity of the yearly category citation-flow network at the category and area partitions, and the within-category self-share of citations.}\label{tab:p9-modularity} 2 \begin{table}[htbp]\centering\caption{Modularity of the yearly category citation-flow network at the category and area partitions, and the within-category self-share of citations.}\label{tab:p9-modularity}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrr} 4 \begin{tabular}{lrrrr}
modified manuscript/tables/tab_p9_openness.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/volume_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/volume_yearly.csv
2 \begin{table}[htbp]\centering\caption{Openness and metadata indicators by publication year: arXiv linkage, preprint share, open access, abstract availability and document type (per cent of eligible works).}\label{tab:p9-openness} 2 \begin{table}[htbp]\centering\caption{Openness and metadata indicators by publication year: arXiv linkage, preprint share, open access, abstract availability and document type (per cent of eligible works).}\label{tab:p9-openness}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_p9_periods.tex +15 −15
@@ -1,24 +1,24 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/volume_monthly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/volume_monthly.csv
2 \begin{table}[htbp]\centering\caption{The eligible corpus by protocol period: completeness, source membership and document type.}\label{tab:p9-periods} 2 \begin{table}[htbp]\centering\caption{The eligible corpus by protocol period: completeness, source membership and document type.}\label{tab:p9-periods}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrr} 4 \begin{tabular}{lrr}
5 \toprule 5 \toprule
6 \textbf{Indicator} & \textbf{Baseline} & \textbf{Post} \\ 6 \textbf{Indicator} & \textbf{Baseline} & \textbf{Post} \\
7 \midrule 7 \midrule
8−Eligible works (count) & 3{,}405{,}453.00 & 3{,}705{,}041.00 \\ 8+Eligible works (count) & 3,405,453 & 3,705,041 \\
9−Month-precision first date (\\%) & 77.43 & 82.78 \\ 9+Month-precision first date (\%) & 77.43 & 82.78 \\
10−DBLP-indexed (\\%) & 54.16 & 47.56 \\ 10+DBLP-indexed (\%) & 54.16 & 47.56 \\
11−OpenAlex primary field CS (\\%) & 72.33 & 77.15 \\ 11+OpenAlex primary field CS (\%) & 72.33 & 77.15 \\
12−arXiv-linked (\\%) & 8.70 & 13.66 \\ 12+arXiv-linked (\%) & 8.70 & 13.66 \\
13−With abstract (\\%) & 73.42 & 77.88 \\ 13+With abstract (\%) & 73.42 & 77.88 \\
14−With at least one reference (\\%) & 71.63 & 58.71 \\ 14+With at least one reference (\%) & 71.63 & 58.71 \\
15−Open access (\\%) & 47.04 & 60.23 \\ 15+Open access (\%) & 47.04 & 60.23 \\
16−Type: article (\\%) & 52.63 & 50.31 \\ 16+Type: article (\%) & 52.63 & 50.31 \\
17−Type: conference paper (\\%) & 31.24 & 29.01 \\ 17+Type: conference paper (\%) & 31.24 & 29.01 \\
18−Type: preprint (\\%) & 5.29 & 11.85 \\ 18+Type: preprint (\%) & 5.29 & 11.85 \\
19−Type: book chapter (\\%) & 5.41 & 4.29 \\ 19+Type: book chapter (\%) & 5.41 & 4.29 \\
20−Type: dissertation (\\%) & 3.99 & 2.74 \\ 20+Type: dissertation (\%) & 3.99 & 2.74 \\
21 \bottomrule 21 \bottomrule
22 \end{tabular} 22 \end{tabular}
23−\begin{tablenotes}\footnotesize\item Baseline: first date 2018-01-01 to 2022-11-29; post: 2022-11-30 to 2026-11-02 (right-truncated, see Part III). Percentages are over all eligible works of the period. The count row is an integer written with two decimals by the table generator.\end{tablenotes} 23+\begin{tablenotes}\footnotesize\item Baseline: first date 2018-01-01 to 2022-11-29; post: 2022-11-30 to 2026-11-02 (right-truncated, see Part III). Percentages are over all eligible works of the period. The first row is a count.\end{tablenotes}
24 \end{threeparttable}\end{table} 24 \end{threeparttable}\end{table}
modified manuscript/tables/tab_p9_total_robustness.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{table}[htbp]\centering\caption{Robustness diagnostics of the structural-change analysis of the total monthly series.}\label{tab:p9-total-robustness} 2 \begin{table}[htbp]\centering\caption{Robustness diagnostics of the structural-change analysis of the total monthly series.}\label{tab:p9-total-robustness}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lr} 4 \begin{tabular}{lr}
modified manuscript/tables/tab_p9_unknown_breaks.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/econometrics/its_results.json 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/its_results.json
2 \begin{longtable}{lrrrrrr} 2 \begin{longtable}{lrrrrrr}
3 \caption{Data-driven break and changepoint dates by category, and whether any falls in the H3 window (October 2022 to June 2023).}\label{tab:p9-unknown-breaks}\\ 3 \caption{Data-driven break and changepoint dates by category, and whether any falls in the H3 window (October 2022 to June 2023).}\label{tab:p9-unknown-breaks}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_reference_coverage.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/reference_coverage_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/reference_coverage_yearly.csv
2 \begin{table}[htbp]\centering\caption{Reference coverage by citing year: declared, resolved, internal and external references of eligible works.}\label{tab:reference_coverage} 2 \begin{table}[htbp]\centering\caption{Reference coverage by citing year: declared, resolved, internal and external references of eligible works.}\label{tab:reference_coverage}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrr} 4 \begin{tabular}{lrrrrrrr}
modified manuscript/tables/tab_security_counts.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_counts_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv
2 \begin{longtable}{lrrrrrrrrrr} 2 \begin{longtable}{lrrrrrrrrrr}
3 \caption{Eligible works in the five security-and-privacy categories by publication year: works carrying the category (all) and works whose primary category it is (prim.).}\label{tab:security_counts}\\ 3 \caption{Eligible works in the five security-and-privacy categories by publication year: works carrying the category (all) and works whose primary category it is (prim.).}\label{tab:security_counts}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_security_landmarks.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./literature_synthesis/landmarks/security_privacy_landmarks.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/literature_synthesis/landmarks/security_privacy_landmarks.csv
2 \begin{table}[htbp]\centering\caption{Composition of the automatically selected landmark set of the security-and-privacy area, by category.}\label{tab:security_landmarks} 2 \begin{table}[htbp]\centering\caption{Composition of the automatically selected landmark set of the security-and-privacy area, by category.}\label{tab:security_landmarks}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_security_llm_cooccurrence.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_cooccurrence_yearly.csv, ./results/graph/category_counts_yearly.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_cooccurrence_yearly.csv, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv
2 \begin{longtable}{lrrrrrrrrrrrrrr} 2 \begin{longtable}{lrrrrrrrrrrrrrr}
3 \caption{Co-classification of security-and-privacy works with the LLM, AI-safety and AI-agents categories, by publication year: number of works carrying both categories and share of the security category's works.}\label{tab:security_llm_cooccurrence}\\ 3 \caption{Co-classification of security-and-privacy works with the LLM, AI-safety and AI-agents categories, by publication year: number of works carrying both categories and share of the security category's works.}\label{tab:security_llm_cooccurrence}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_security_summary.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/graph/category_counts_yearly.csv, ./results/econometrics/breaks.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/graph/category_counts_yearly.csv, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv
2 \begin{table}[htbp]\centering\caption{Growth and structural-change summary for the five security-and-privacy categories (multi-label counts).}\label{tab:security_summary} 2 \begin{table}[htbp]\centering\caption{Growth and structural-change summary for the five security-and-privacy categories (multi-label counts).}\label{tab:security_summary}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrrrr}
modified manuscript/tables/tab_source_types.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/top_sources_by_period.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_sources_by_period.csv
2 \begin{table}[htbp]\centering\caption{Distribution of works over source types among the listed sources, by period (per cent).}\label{tab:source_types} 2 \begin{table}[htbp]\centering\caption{Distribution of works over source types among the listed sources, by period (per cent).}\label{tab:source_types}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrr} 4 \begin{tabular}{lrrrrrrr}
modified manuscript/tables/tab_systems_categories.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/cagr_by_category.csv, ./results/econometrics/breaks.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/cagr_by_category.csv, /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/econometrics/breaks.csv
2 \begin{table}[htbp]\centering\caption{The systems and infrastructure categories of Part VI: primary-category output, compound annual growth, and structural-change estimates of the monthly log-volume series.}\label{tab:systems_categories} 2 \begin{table}[htbp]\centering\caption{The systems and infrastructure categories of Part VI: primary-category output, compound annual growth, and structural-change estimates of the monthly log-volume series.}\label{tab:systems_categories}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrrrrrr} 4 \begin{tabular}{lrrrrrrrrr}
modified manuscript/tables/tab_top_cited.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/top_cited_by_year.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_cited_by_year.csv
2 \begin{longtable}{lrrrr} 2 \begin{longtable}{lrrrr}
3 \caption{The ten most cited works of each publication year by 36-month window citations (c36), with total citations to date.}\label{tab:top_cited}\\ 3 \caption{The ten most cited works of each publication year by 36-month window citations (c36), with total citations to date.}\label{tab:top_cited}\\
4 \toprule 4 \toprule
modified manuscript/tables/tab_top_institutions.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/top_institutions_by_period.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_institutions_by_period.csv
2 \begin{table}[htbp]\centering\caption{The twenty-five institutions with the largest fractional output in the post period, with their baseline output.}\label{tab:top_institutions} 2 \begin{table}[htbp]\centering\caption{The twenty-five institutions with the largest fractional output in the post period, with their baseline output.}\label{tab:top_institutions}
3 \begin{threeparttable}\small 3 \begin{threeparttable}\small
4 \begin{tabular}{lrrrrr} 4 \begin{tabular}{lrrrrr}
modified manuscript/tables/tab_top_sources.tex +1 −1
@@ -1,4 +1,4 @@
1−% generated by tables/make_tables.py from ./results/bibliometrics/top_sources_by_period.csv 1+% generated by tables/make_tables.py from /Users/simon-pierreboucher/Desktop/UQO/Recherche/Recherche-UQO/informatique/wp44/results/bibliometrics/top_sources_by_period.csv
2 \begin{longtable}{lrrrrrr} 2 \begin{longtable}{lrrrrrr}
3 \caption{The twenty largest sources (venues) by period, with OpenAlex source type.}\label{tab:top_sources}\\ 3 \caption{The twenty largest sources (venues) by period, with OpenAlex source type.}\label{tab:top_sources}\\
4 \toprule 4 \toprule
modified results/bibliometrics/cagr_by_category.csv +40 −40
@@ -1,41 +1,41 @@
1 category,n_2018,n_2022,n_2023,n_2025,cagr_baseline,cagr_post,share_2018,share_2022,share_2025 1 category,n_2018,n_2022,n_2023,n_2025,cagr_baseline,cagr_post,share_2018,share_2022,share_2025
2−ai_accelerators,818,1421,1599,1906,0.14804838919248775,0.09178523385922643,0.001637437670274501,0.002535539484648477,0.0021347609874109584 2+ai_accelerators,409,710,799,953,0.14804838919248775,0.09178523385922643,0.002409752131363896,0.0027899579837432427,0.002218169672523209
3−ai_agents,5436,4521,4991,8819,-0.045032119445381436,0.3292782120448039,0.010881554004415877,0.008066976784022354,0.00987746964741723 3+ai_agents,2718,2260,2495,4409,-0.045032119445381436,0.3292782120448039,0.016013951816741003,0.00887642508409796,0.010263398920242486
4−ai_safety,1052,2954,4015,10885,0.2944899640897396,0.6465368010192893,0.0021058489353652505,0.005270924445919495,0.01219143407553425 4+ai_safety,526,1477,2007,5442,0.2944899640897396,0.6465368010192893,0.0030990944281110255,0.005799814133692849,0.01266777381186523
5−algorithms,29148,28202,28531,33795,-0.008214426291722088,0.08834786213446155,0.058347228866945176,0.050321804747400674,0.03785112674163344 5+algorithms,9716,9400,9510,11265,-0.008214426291722088,0.08834786213446155,0.05724486970252229,0.03691409573423735,0.02622002241445325
6−applied_cryptography,10782,10299,11219,14612,-0.01139242765304549,0.14124203040296623,0.021582949829950696,0.01837686217621018,0.01636575422248107 6+applied_cryptography,3594,5149,3739,7306,0.09407412599520226,0.3977303237525569,0.0211751813205913,0.020220814408565557,0.01700519163426502
7−autonomous_scientific_discovery,86,236,400,764,0.28707361667658393,0.38202749610852527,0.00017215114870856613,0.00042110296859749515,0.000855696429371444 7+autonomous_scientific_discovery,43,118,200,382,0.28707361667658393,0.38202749610852527,0.0002533480235908253,0.0004633568502205526,0.0008891299212002788
8−blockchain,6441,11770,14349,18012,0.1626681909797465,0.12039239265851642,0.012893320335254354,0.02100161839149372,0.02017382733748488 8+blockchain,2147,5885,7174,9006,0.2867038655679659,0.12039239265851642,0.012649725736034926,0.02310894121650807,0.020962052540130133
9−cloud_computing,9396,7103,7643,10974,-0.06755182852386099,0.19825858205350233,0.01880851387518241,0.012674128754016984,0.012291115989426997 9+cloud_computing,3132,3551,3821,3658,0.03192353243026824,-0.0216259646860103,0.01845316302061546,0.013945863165748243,0.00851423364332623
10−compilers,725,642,692,767,-0.029938529449611395,0.05279699035007379,0.0014512742187640748,0.001145542821354203,0.0008590564938846826 10+compilers,241,321,346,383,0.073549490456021,0.05279699035007379,0.0014238551713437854,0.001260487702718622,0.0008926212690583952
11−computational_biology,12968,19536,20576,28154,0.10787474203944436,0.1697406366696017,0.025958791819217272,0.034858760993731634,0.03153308543524036 11+computational_biology,4322,9768,10288,14077,0.22606494630771068,0.1697406366696017,0.025468350154463738,0.03835652299113863,0.032765135865801896
12−computational_complexity,3837,3541,3499,4209,-0.019870339595214892,0.0967748714322898,0.007680743692962421,0.006318328863575129,0.004714170512073831 12+computational_complexity,1279,1770,1166,2104,0.08469177413199191,0.34326939885786634,0.007535630748201524,0.006952316129792274,0.004898361044937139
13−computer_architecture,19554,15174,16369,20209,-0.06143119143120024,0.1111209535203932,0.03914236699822444,0.027075493413128777,0.02263451458267999 13+computer_architecture,6518,7587,8184,6736,0.03869713083776305,-0.0927735404534441,0.03840284692476742,0.029792274767994346,0.015679255303260413
14−computer_graphics,10198,8287,9297,12440,-0.050553314316049036,0.15674804638240403,0.020413923424766944,0.014786780935455265,0.013933067514896286 14+computer_graphics,3399,4143,3099,6220,0.05073548076582268,0.4167212372990894,0.02002824143084679,0.016270500922786948,0.014477455784988833
15−computer_networks,58744,49141,51881,59555,-0.04364282894058291,0.07140814904065751,0.1175912451132094,0.08768398720275215,0.06670288069530934 15+computer_networks,19581,16380,17293,29777,-0.04364282894058291,0.3122016357047006,0.11536958370402665,0.06432152253301744,0.06930907389670499
16−computer_vision,69738,98704,114261,150757,0.09072763371648507,0.1486553322088684,0.13959856754230215,0.17612096361206425,0.16885108194077325 16+computer_vision,23246,49352,57130,75378,0.20708855155154882,0.1486553322088684,0.13696111991610055,0.19379311247529418,0.17544837634867858
17−cybersecurity,29440,31554,36481,59044,0.017487667119059358,0.2721974435411547,0.05893174206953705,0.056302894369175266,0.06613054970655437 17+cybersecurity,9813,15777,12160,29522,0.12603520471872232,0.5581172943745198,0.057818339647394544,0.06195238157567507,0.06871438097820584
18−database_systems,16265,13625,14487,16681,-0.043311373704336575,0.07305457648307523,0.032558586438893346,0.02431155909805454,0.01868307871511133 18+database_systems,5421,6812,7243,8340,0.05875000550434839,0.07305457648307523,0.03194345429228506,0.02675100459430097,0.01941305787374588
19−deep_learning_architectures,11992,22270,25607,34141,0.16736629336390108,0.15467235481054153,0.02400507645712936,0.03973713182485685,0.0382386541821603 19+deep_learning_architectures,3997,7423,12803,17070,0.16736629336390108,0.15467235481054153,0.023551546503109897,0.029149596198903123,0.039732702407982615
20−distributed_computing,11519,9379,9798,12550,-0.050083482537713886,0.13175688541453123,0.023058245139232245,0.01673527433252503,0.014056269880381703 20+distributed_computing,3839,4689,3266,4183,0.0512554351002068,0.13175688541453123,0.022622603749943536,0.018414508043298996,0.009736981248746508
21−edge_computing,5185,8554,9120,11554,0.1333267023064395,0.1255602893280665,0.010379112861092039,0.015263198277046497,0.01294072846198647 21+edge_computing,1728,4277,3040,5777,0.25423217055612524,0.3785241917965825,0.010183019397817279,0.016794722444011048,0.01344634438422516
22−federated_learning,143,974,1497,2544,0.6154954153196714,0.303609910055608,0.00028625132866656923,0.001737941912771018,0.002849334707226379 22+federated_learning,71,487,748,1272,0.6154954153196714,0.303609910055608,0.0004212647369010234,0.0019123286954017722,0.002960662983682604
23−formal_verification,5307,3995,4385,4623,-0.06853417817252339,0.026779402163364052,0.010623327281353028,0.0071284167777414965,0.0051778594149007665 23+formal_verification,1769,1997,1461,2311,0.030836383835269654,0.2575428068501022,0.010422619854236509,0.007843689053521643,0.005380167049357185
24−foundation_models,46,350,571,1854,0.6608390716251988,0.8019254150494273,9.208084698365164e-05,0.0006245171144454377,0.0020765198691814883 24+foundation_models,23,175,285,927,0.6608390716251988,0.8019254150494273,0.00013551173354858095,0.0006871817693948874,0.00215765297631586
25−generative_ai,766,2160,3646,12743,0.29585451247314687,0.8695087811589706,0.0015333462780321122,0.0038541627634347013,0.01427243403073339 25+generative_ai,255,1080,1823,6371,0.43409875951610655,0.8695087811589706,0.0015043766362059857,0.004240893205408448,0.014830081918658578
26−hci,34225,34801,40587,58148,0.004181154443647861,0.19694422895345198,0.06851015191337995,0.06209662885661622,0.06512701043860042 26+hci,11408,17400,20293,29074,0.11130912772658119,0.19694422895345198,0.06721578377826352,0.06832746501917564,0.06767163175124845
27−information_retrieval,12095,11243,12568,17379,-0.018095875045832766,0.1759241392022266,0.024211257484071016,0.02006127405060016,0.019464853725191522 27+information_retrieval,4031,5621,6284,8689,0.08665554196594227,0.1759241392022266,0.023753832134349083,0.022074241809447767,0.0202253781420676
28−llm,1635,3266,10389,45158,0.1888432761609986,1.0848771832846835,0.003272873583005879,0.005827636845082284,0.05057793109627705 28+llm,545,1088,5194,22579,0.1888432761609986,1.0848771832846835,0.0032110389036511576,0.004274925064464195,0.052554095525604964
29−ml_theory,13534,20131,22893,32798,0.10435777777595234,0.19693985094098543,0.02709178658862481,0.035920440088288874,0.03673446530173379 29+ml_theory,4511,10065,11446,16399,0.22217278554507214,0.19693985094098543,0.026579939157195573,0.039524731999109934,0.038169742350165896
30−multimodal_ai,470,1278,1932,6749,0.2841278026110756,0.8690294865336103,0.000940826045267745,0.002280379635032198,0.007559025133282559 30+multimodal_ai,156,639,966,3374,0.421121021734181,0.8690294865336103,0.000923050938664247,0.0025091951465333316,0.007854368898142254
31−nlp,23372,29678,33965,42914,0.06153590511416507,0.12404497631921241,0.04678507729786753,0.052955482635747717,0.04806460284037453 31+nlp,7790,9892,16982,21457,0.06153590511416507,0.12404497631921241,0.045901162847788904,0.038846058194478984,0.049942567327733986
32−operating_systems,1556,1083,1173,1363,-0.0866132316020718,0.07795075704417309,0.003114734737099173,0.0019324343855554546,0.001526589310514762 32+operating_systems,778,541,586,681,-0.0866132316020718,0.07795075704417309,0.004583831682643304,0.00212633673215618,0.0015862357102041625
33−privacy_preserving_computation,3105,4376,5010,8817,0.08956696494067806,0.3266047789454696,0.006215457171396486,0.007808248265180673,0.009875229604408405 33+privacy_preserving_computation,1035,2188,2505,4408,0.20580406040260257,0.3266047789454696,0.006098028009686143,0.008591735493920076,0.010261071355003742
34−programming_languages,4390,3664,3876,4193,-0.044187433279510024,0.04008910470076277,0.008787715614309364,0.006537802020937383,0.004696250168003226 34+programming_languages,1463,1832,1938,2096,0.057780486211923554,0.04008910470076277,0.008621688554757541,0.007193811437322478,0.0048797405230271845
35−quantum_computing,6224,9216,10544,17471,0.10310847324827721,0.2872300901577727,0.012458938948396692,0.016444427790654727,0.01956789570359751 35+quantum_computing,3112,4608,5272,8735,0.10310847324827721,0.2872300901577727,0.018335326730573216,0.018094477676409378,0.020332446143049833
36−reinforcement_learning,3296,7017,8422,13210,0.20792805552654703,0.25240210895212,0.006597792862132953,0.012520675977324676,0.014795484073294207 36+reinforcement_learning,1648,3508,4211,6605,0.20792805552654703,0.25240210895212,0.009709710299480931,0.013777012788125499,0.015373568401905345
37−robotics,17496,23103,25759,33464,0.07196943376447251,0.1397889079167418,0.035022749974477595,0.04122348255723699,0.03748039962367278 37+robotics,5832,11551,12879,16732,0.18632919081853316,0.1397889079167418,0.034361062176318444,0.04535988690951452,0.038944821574667714
38−scientific_computing,16882,19932,20423,23191,0.04239362249738088,0.06561412676886369,0.033793670843000154,0.03556535750036133,0.02597441870883921 38+scientific_computing,5627,6644,6807,7730,0.04239362249738088,0.06561412676886369,0.033155204141552805,0.026089346719197895,0.017992855150572133
39−software_engineering,29016,23725,26245,36381,-0.04908396762848111,0.17737287914735167,0.05808299687125296,0.04233333868633717,0.04074750235204516 39+software_engineering,9672,11862,13122,18190,0.05236158017872561,0.17737287914735167,0.05698562986442935,0.04658110708255343,0.042339575475376104
40−speech_audio,12096,12698,14688,18373,0.012216438797881812,0.11842973684211522,0.024213259241614136,0.02265748091208048,0.020578155100577932 40+speech_audio,4032,6349,7344,9186,0.12020163164106301,0.11842973684211522,0.023755796072516454,0.024930954593646512,0.02138217806572346
41−sustainable_computing,583,830,1078,1639,0.09232628514238472,0.23304832154514776,0.0011670246476406284,0.001480997728542038,0.0018357152457327182 41+sustainable_computing,291,276,539,819,-0.012971780149642509,0.23304832154514776,0.0017174639273657107,0.0010864016544719173,0.00190743971315086
added results/bibliometrics/category_actors_yearly.csv +731 −0
@@ -0,0 +1,731 @@
1+year,period,primary_category,n,share_cn,share_us,share_company,share_intl,share_arxiv,share_oa,mean_authors
2+2018,baseline,ai_accelerators,816,0.2169,0.2598,0.0686,0.2059,0.0846,0.3407,4.152
3+2018,pre,ai_accelerators,2,0.0,0.5,0.5,0.0,1.0,1.0,3.5
4+2018,pre,ai_agents,15,0.2,0.2667,0.1333,0.2,1.0,0.9333,2.8
5+2018,baseline,ai_agents,5421,0.1443,0.1136,0.0249,0.162,0.0644,0.3935,2.779
6+2018,baseline,ai_safety,1049,0.0305,0.2574,0.0362,0.1163,0.1153,0.4471,2.359
7+2018,pre,ai_safety,3,0.0,0.6667,0.3333,0.6667,1.0,1.0,5.667
8+2018,baseline,algorithms,29011,0.1579,0.1516,0.0271,0.2131,0.09,0.4858,2.763
9+2018,post,algorithms,8,0.0,0.375,0.125,0.5,1.0,1.0,3.375
10+2018,pre,algorithms,129,0.031,0.3488,0.0465,0.4031,0.9612,0.9457,2.829
11+2018,post,applied_cryptography,1,0.0,1.0,0.0,0.0,1.0,1.0,4.0
12+2018,baseline,applied_cryptography,10764,0.1455,0.085,0.0334,0.1335,0.0295,0.3748,2.785
13+2018,pre,applied_cryptography,17,0.2353,0.2353,0.0,0.3529,0.9412,0.8235,2.647
14+2018,baseline,autonomous_scientific_discovery,86,0.0465,0.2558,0.093,0.1744,0.0814,0.4535,3.279
15+2018,baseline,blockchain,6435,0.06,0.0783,0.0242,0.0779,0.0292,0.3711,2.276
16+2018,pre,blockchain,6,0.1667,0.1667,0.1667,0.6667,1.0,1.0,4.0
17+2018,baseline,cloud_computing,9395,0.1214,0.0919,0.0467,0.1236,0.0141,0.3222,2.815
18+2018,pre,cloud_computing,1,0.0,0.0,0.0,1.0,1.0,1.0,3.0
19+2018,baseline,compilers,725,0.08,0.1862,0.0538,0.1269,0.029,0.4193,3.036
20+2018,baseline,computational_biology,12957,0.152,0.2667,0.0485,0.2222,0.0203,0.5632,4.615
21+2018,pre,computational_biology,11,0.0,0.3636,0.0909,0.3636,1.0,1.0,5.727
22+2018,baseline,computational_complexity,3822,0.0905,0.1434,0.016,0.1617,0.0858,0.505,2.388
23+2018,pre,computational_complexity,15,0.0,0.4667,0.0667,0.4,1.0,1.0,2.8
24+2018,pre,computer_architecture,12,0.0,0.25,0.1667,0.1667,1.0,1.0,4.333
25+2018,baseline,computer_architecture,19542,0.1087,0.1964,0.0902,0.1498,0.0224,0.2796,3.477
26+2018,baseline,computer_graphics,10187,0.1091,0.1343,0.0433,0.1222,0.0295,0.3746,2.787
27+2018,pre,computer_graphics,11,0.1818,0.0909,0.0909,0.1818,1.0,1.0,3.0
28+2018,post,computer_networks,1,0.0,0.0,0.0,0.0,1.0,1.0,1.0
29+2018,baseline,computer_networks,58594,0.2048,0.1244,0.0651,0.1942,0.0558,0.3386,3.309
30+2018,pre,computer_networks,149,0.2282,0.3826,0.1007,0.3624,0.9933,0.9732,3.383
31+2018,pre,computer_vision,148,0.2635,0.3311,0.0811,0.3919,1.0,0.9662,4.284
32+2018,baseline,computer_vision,69585,0.2526,0.1157,0.0496,0.1485,0.0931,0.3713,3.558
33+2018,post,computer_vision,5,0.4,0.2,0.0,0.0,1.0,1.0,4.8
34+2018,baseline,cybersecurity,29406,0.098,0.1238,0.0334,0.1028,0.0443,0.368,2.755
35+2018,post,cybersecurity,2,0.0,0.0,0.0,0.0,1.0,1.0,1.5
36+2018,pre,cybersecurity,32,0.1563,0.5313,0.0938,0.3438,1.0,1.0,3.594
37+2018,pre,database_systems,9,0.1111,0.3333,0.0,0.5556,1.0,1.0,4.556
38+2018,baseline,database_systems,16256,0.1085,0.1054,0.0369,0.122,0.028,0.4086,2.862
39+2018,pre,deep_learning_architectures,41,0.1951,0.439,0.1463,0.4878,1.0,0.9512,4.39
40+2018,baseline,deep_learning_architectures,11950,0.225,0.1468,0.0507,0.1544,0.1396,0.4147,3.261
41+2018,post,deep_learning_architectures,1,0.0,1.0,0.0,0.0,1.0,1.0,2.0
42+2018,baseline,distributed_computing,11489,0.1597,0.1834,0.0498,0.1591,0.0638,0.3925,3.251
43+2018,post,distributed_computing,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
44+2018,pre,distributed_computing,29,0.1034,0.4483,0.0345,0.4483,1.0,1.0,3.207
45+2018,pre,edge_computing,3,0.0,1.0,0.0,0.3333,1.0,1.0,3.333
46+2018,baseline,edge_computing,5182,0.115,0.1324,0.0556,0.191,0.0152,0.3203,3.361
47+2018,baseline,federated_learning,143,0.1399,0.2448,0.0629,0.1818,0.028,0.4336,3.413
48+2018,baseline,formal_verification,5278,0.0892,0.1408,0.0396,0.2363,0.1465,0.5731,2.672
49+2018,pre,formal_verification,29,0.1034,0.2069,0.0,0.3793,1.0,1.0,2.517
50+2018,baseline,foundation_models,46,0.2174,0.1957,0.0435,0.1087,0.2391,0.4348,3.891
51+2018,baseline,generative_ai,766,0.1841,0.2272,0.0927,0.1253,0.2911,0.4739,3.358
52+2018,pre,hci,12,0.0833,0.25,0.0,0.25,1.0,1.0,4.333
53+2018,post,hci,4,0.25,0.0,0.0,0.0,1.0,1.0,3.75
54+2018,baseline,hci,34209,0.0517,0.1338,0.035,0.1137,0.0169,0.376,2.968
55+2018,pre,information_retrieval,16,0.0625,0.3125,0.0625,0.375,1.0,1.0,3.563
56+2018,baseline,information_retrieval,12077,0.1673,0.124,0.054,0.1456,0.0745,0.3958,2.937
57+2018,post,information_retrieval,2,0.0,0.0,0.0,0.0,1.0,1.0,3.0
58+2018,baseline,llm,1633,0.1249,0.0802,0.0374,0.0986,0.0245,0.2817,2.994
59+2018,pre,llm,2,0.0,0.5,0.0,0.5,1.0,1.0,5.0
60+2018,post,ml_theory,1,0.0,1.0,1.0,1.0,1.0,1.0,5.0
61+2018,baseline,ml_theory,13468,0.1529,0.2073,0.0595,0.161,0.2493,0.5418,3.019
62+2018,pre,ml_theory,65,0.1385,0.4769,0.0923,0.3538,1.0,0.9538,3.385
63+2018,pre,multimodal_ai,1,0.0,0.0,0.0,1.0,1.0,1.0,3.0
64+2018,baseline,multimodal_ai,469,0.3369,0.2132,0.0959,0.1812,0.1279,0.4051,3.62
65+2018,baseline,nlp,23347,0.1141,0.1332,0.0572,0.1247,0.1158,0.5092,3.04
66+2018,pre,nlp,22,0.0909,0.4545,0.1364,0.5,1.0,1.0,4.182
67+2018,post,nlp,3,0.0,0.3333,0.0,0.6667,1.0,1.0,2.333
68+2018,baseline,operating_systems,1554,0.1338,0.1577,0.0727,0.1178,0.0161,0.3333,3.027
69+2018,pre,operating_systems,2,0.0,0.5,0.5,0.5,1.0,1.0,4.0
70+2018,baseline,privacy_preserving_computation,3098,0.1724,0.2198,0.0462,0.1756,0.0594,0.4032,3.255
71+2018,pre,privacy_preserving_computation,7,0.0,0.1429,0.2857,0.5714,1.0,0.8571,3.571
72+2018,baseline,programming_languages,4383,0.0418,0.1476,0.0331,0.141,0.102,0.5086,2.277
73+2018,post,programming_languages,1,0.0,0.0,0.0,1.0,1.0,1.0,5.0
74+2018,pre,programming_languages,6,0.0,0.3333,0.1667,0.6667,1.0,1.0,3.667
75+2018,baseline,quantum_computing,6187,0.2017,0.1474,0.0391,0.2022,0.0438,0.5495,3.594
76+2018,pre,quantum_computing,37,0.1081,0.1892,0.1081,0.4595,1.0,1.0,2.946
77+2018,baseline,reinforcement_learning,3289,0.1538,0.2195,0.09,0.145,0.2277,0.4865,3.212
78+2018,pre,reinforcement_learning,7,0.0,0.2857,0.0,0.2857,1.0,1.0,2.571
79+2018,pre,robotics,58,0.1034,0.569,0.0517,0.2586,1.0,0.931,3.793
80+2018,baseline,robotics,17437,0.1746,0.1812,0.0468,0.1681,0.0818,0.3658,3.657
81+2018,post,robotics,1,0.0,0.0,0.0,0.0,1.0,1.0,5.0
82+2018,pre,scientific_computing,467,0.1178,0.3405,0.0086,0.3426,1.0,0.9529,2.835
83+2018,baseline,scientific_computing,16414,0.226,0.1732,0.0344,0.2204,0.073,0.4828,3.055
84+2018,post,scientific_computing,1,0.0,1.0,0.0,1.0,1.0,1.0,3.0
85+2018,pre,software_engineering,16,0.125,0.3125,0.0,0.4375,1.0,1.0,5.5
86+2018,baseline,software_engineering,29000,0.0525,0.0828,0.0377,0.1001,0.0277,0.3541,2.471
87+2018,post,speech_audio,1,0.0,0.0,0.0,0.0,1.0,1.0,4.0
88+2018,baseline,speech_audio,12081,0.1153,0.0959,0.0561,0.1069,0.0699,0.351,3.024
89+2018,pre,speech_audio,14,0.2857,0.3571,0.1429,0.4286,1.0,0.9286,3.714
90+2018,pre,sustainable_computing,2,0.0,0.5,0.0,0.5,1.0,1.0,2.5
91+2018,baseline,sustainable_computing,581,0.1515,0.1549,0.062,0.2375,0.0224,0.4131,3.444
92+2019,baseline,ai_accelerators,992,0.2339,0.253,0.0665,0.2177,0.0938,0.3538,4.363
93+2019,baseline,ai_agents,5503,0.1703,0.1188,0.0305,0.1652,0.0836,0.4136,2.856
94+2019,post,ai_agents,2,0.0,0.0,0.0,0.0,1.0,1.0,2.0
95+2019,pre,ai_agents,1,0.0,0.0,0.0,0.0,1.0,0.0,1.0
96+2019,pre,ai_safety,1,0.0,0.0,0.0,0.0,1.0,1.0,1.0
97+2019,baseline,ai_safety,1478,0.0359,0.2659,0.0514,0.1238,0.1543,0.4926,2.565
98+2019,pre,algorithms,35,0.0571,0.2,0.0286,0.3714,1.0,0.9429,2.743
99+2019,baseline,algorithms,29204,0.1835,0.1542,0.031,0.2234,0.1227,0.5259,2.852
100+2019,post,algorithms,13,0.0,0.3846,0.0769,0.3846,1.0,1.0,3.385
101+2019,baseline,applied_cryptography,11352,0.1684,0.0847,0.0344,0.1373,0.0419,0.4066,2.885
102+2019,pre,applied_cryptography,3,0.0,0.0,0.0,0.0,1.0,1.0,3.0
103+2019,baseline,autonomous_scientific_discovery,114,0.0263,0.3333,0.0702,0.1842,0.114,0.5439,4.149
104+2019,baseline,blockchain,9045,0.0826,0.0871,0.0238,0.0905,0.0305,0.3787,2.545
105+2019,post,blockchain,2,0.0,0.0,0.0,0.0,1.0,1.0,2.5
106+2019,pre,cloud_computing,1,0.0,1.0,0.0,1.0,1.0,1.0,3.0
107+2019,baseline,cloud_computing,8755,0.136,0.0899,0.0474,0.1261,0.0162,0.3676,2.871
108+2019,baseline,compilers,689,0.0827,0.1872,0.0639,0.1176,0.0392,0.4673,2.955
109+2019,pre,computational_biology,1,0.0,1.0,0.0,0.0,1.0,1.0,5.0
110+2019,baseline,computational_biology,14875,0.1787,0.261,0.0529,0.2223,0.026,0.5882,4.68
111+2019,pre,computational_complexity,10,0.0,0.3,0.0,0.2,1.0,1.0,2.6
112+2019,baseline,computational_complexity,3941,0.1058,0.1373,0.018,0.1913,0.1056,0.5227,2.486
113+2019,post,computational_complexity,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
114+2019,post,computer_architecture,1,0.0,0.0,0.0,1.0,1.0,1.0,15.0
115+2019,baseline,computer_architecture,18844,0.1183,0.2029,0.0916,0.1516,0.0272,0.3022,3.567
116+2019,pre,computer_architecture,1,0.0,0.0,0.0,0.0,1.0,1.0,3.0
117+2019,pre,computer_graphics,2,0.5,1.0,0.0,0.5,1.0,1.0,4.0
118+2019,baseline,computer_graphics,9796,0.1134,0.1365,0.0475,0.1241,0.0481,0.4282,2.898
119+2019,baseline,computer_networks,58278,0.2277,0.1205,0.0616,0.2006,0.0655,0.3664,3.394
120+2019,pre,computer_networks,18,0.2778,0.3333,0.2222,0.3889,1.0,0.8333,3.111
121+2019,post,computer_networks,3,0.0,0.0,0.0,0.0,1.0,1.0,2.0
122+2019,baseline,computer_vision,81632,0.2773,0.1183,0.0604,0.1579,0.1216,0.415,3.712
123+2019,post,computer_vision,11,0.3636,0.0,0.2727,0.3636,1.0,1.0,7.182
124+2019,pre,computer_vision,27,0.2963,0.2963,0.0,0.2963,1.0,1.0,4.37
125+2019,baseline,cybersecurity,31551,0.103,0.128,0.0366,0.117,0.0553,0.4075,2.834
126+2019,pre,cybersecurity,3,0.0,0.3333,0.0,0.3333,1.0,1.0,5.0
127+2019,post,cybersecurity,7,0.0,0.4286,0.0,0.2857,1.0,1.0,3.714
128+2019,post,database_systems,1,1.0,0.0,0.0,0.0,1.0,1.0,4.0
129+2019,pre,database_systems,4,0.0,0.25,0.5,0.0,1.0,1.0,2.75
130+2019,baseline,database_systems,16574,0.1078,0.1104,0.0404,0.1234,0.0336,0.4532,2.934
131+2019,post,deep_learning_architectures,7,0.0,0.2857,0.1429,0.2857,1.0,1.0,3.0
132+2019,baseline,deep_learning_architectures,15350,0.2506,0.1516,0.0593,0.1745,0.1738,0.4552,3.556
133+2019,pre,deep_learning_architectures,6,0.1667,0.5,0.0,0.3333,1.0,1.0,3.167
134+2019,pre,distributed_computing,5,0.2,0.2,0.0,0.4,1.0,1.0,3.0
135+2019,baseline,distributed_computing,11487,0.1791,0.1977,0.0504,0.1726,0.0868,0.4341,3.368
136+2019,post,distributed_computing,3,0.0,0.0,0.0,0.3333,1.0,1.0,2.0
137+2019,baseline,edge_computing,6649,0.1396,0.1148,0.0486,0.1785,0.0203,0.3443,3.359
138+2019,post,edge_computing,1,0.0,0.0,0.0,0.0,1.0,1.0,3.0
139+2019,baseline,federated_learning,216,0.1806,0.25,0.0648,0.1713,0.0648,0.4907,3.472
140+2019,post,formal_verification,7,0.0,0.0,0.0,0.7143,1.0,0.8571,2.571
141+2019,baseline,formal_verification,4788,0.1005,0.1316,0.0357,0.2435,0.1976,0.5777,2.776
142+2019,pre,formal_verification,7,0.1429,0.2857,0.0,0.7143,1.0,1.0,2.571
143+2019,baseline,foundation_models,83,0.3012,0.2651,0.1205,0.2289,0.1928,0.4458,4.434
144+2019,baseline,generative_ai,1266,0.237,0.1943,0.0853,0.1469,0.2488,0.466,3.575
145+2019,post,hci,2,0.5,0.0,0.0,0.5,1.0,1.0,4.0
146+2019,baseline,hci,37208,0.0587,0.1417,0.0349,0.1171,0.0223,0.3994,3.066
147+2019,pre,hci,2,0.0,1.0,0.0,0.0,1.0,1.0,3.0
148+2019,post,information_retrieval,2,0.0,0.0,0.0,1.0,1.0,1.0,2.0
149+2019,baseline,information_retrieval,12298,0.1943,0.1383,0.0692,0.1511,0.0916,0.436,3.35
150+2019,pre,information_retrieval,4,0.5,0.25,0.25,0.25,1.0,1.0,4.25
151+2019,baseline,llm,1857,0.1341,0.0964,0.0598,0.0894,0.0549,0.3409,3.073
152+2019,baseline,ml_theory,17024,0.1535,0.2274,0.0773,0.1634,0.3046,0.5953,3.149
153+2019,pre,ml_theory,21,0.1905,0.4286,0.0952,0.4286,1.0,0.9524,3.143
154+2019,post,ml_theory,11,0.0909,0.3636,0.3636,0.3636,1.0,1.0,3.455
155+2019,baseline,multimodal_ai,632,0.394,0.1883,0.1108,0.1978,0.1883,0.4225,3.908
156+2019,baseline,nlp,25725,0.1328,0.1341,0.0681,0.1228,0.136,0.5435,3.154
157+2019,pre,nlp,2,0.0,0.0,0.5,0.0,1.0,1.0,2.5
158+2019,post,nlp,11,0.0,0.3636,0.0909,0.3636,1.0,1.0,3.364
159+2019,baseline,operating_systems,1578,0.1394,0.1584,0.0735,0.1255,0.0342,0.3878,2.999
160+2019,pre,operating_systems,1,0.0,1.0,1.0,0.0,1.0,1.0,5.0
161+2019,baseline,privacy_preserving_computation,3256,0.1895,0.1987,0.0501,0.1827,0.0934,0.4487,3.325
162+2019,pre,programming_languages,3,0.0,0.3333,0.0,0.3333,1.0,1.0,2.667
163+2019,baseline,programming_languages,4471,0.0512,0.1436,0.0282,0.1454,0.1118,0.5328,2.365
164+2019,pre,quantum_computing,6,0.1667,0.1667,0.1667,0.5,1.0,1.0,3.667
165+2019,baseline,quantum_computing,7473,0.195,0.1551,0.0541,0.2061,0.055,0.5535,3.755
166+2019,post,quantum_computing,3,1.0,0.0,0.0,0.3333,1.0,1.0,5.0
167+2019,baseline,reinforcement_learning,4735,0.1757,0.237,0.1012,0.1415,0.2769,0.5166,3.289
168+2019,pre,reinforcement_learning,3,0.0,0.3333,0.3333,0.6667,1.0,1.0,4.0
169+2019,post,reinforcement_learning,3,0.0,0.3333,0.3333,0.0,1.0,1.0,1.333
170+2019,post,robotics,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
171+2019,pre,robotics,9,0.1111,0.5556,0.1111,0.4444,1.0,0.6667,2.778
172+2019,baseline,robotics,20045,0.2113,0.1761,0.0529,0.1707,0.1106,0.3911,3.786
173+2019,post,scientific_computing,3,0.0,0.0,0.0,0.0,1.0,1.0,2.333
174+2019,pre,scientific_computing,102,0.1373,0.2353,0.0196,0.3529,1.0,0.951,2.598
175+2019,baseline,scientific_computing,18219,0.2475,0.1713,0.0354,0.2323,0.1239,0.5402,3.135
176+2019,pre,software_engineering,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
177+2019,post,software_engineering,4,0.0,0.0,0.0,0.25,1.0,1.0,3.75
178+2019,baseline,software_engineering,28930,0.0577,0.0777,0.037,0.1046,0.0358,0.3823,2.566
179+2019,baseline,speech_audio,12780,0.1381,0.1002,0.065,0.1126,0.0982,0.4074,3.1
180+2019,pre,speech_audio,3,0.0,0.3333,0.0,0.3333,1.0,1.0,1.667
181+2019,baseline,sustainable_computing,667,0.1664,0.1739,0.0405,0.1979,0.027,0.4018,3.261
182+2020,baseline,ai_accelerators,1085,0.2507,0.2406,0.0608,0.2461,0.1134,0.4101,4.532
183+2020,post,ai_agents,3,0.0,0.0,0.0,0.0,1.0,1.0,4.0
184+2020,baseline,ai_agents,5126,0.1851,0.1092,0.0308,0.1773,0.1178,0.4992,2.967
185+2020,post,ai_safety,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
186+2020,baseline,ai_safety,2153,0.0376,0.2606,0.0646,0.1319,0.1709,0.5625,2.7
187+2020,baseline,algorithms,29342,0.1856,0.1522,0.0313,0.2337,0.1347,0.554,2.888
188+2020,post,algorithms,23,0.0435,0.2174,0.0435,0.3913,1.0,1.0,2.913
189+2020,pre,algorithms,11,0.0,0.2727,0.0,0.1818,0.9091,1.0,2.636
190+2020,pre,applied_cryptography,1,1.0,0.0,0.0,0.0,1.0,1.0,1.0
191+2020,post,applied_cryptography,2,0.0,0.0,0.0,0.0,1.0,1.0,2.5
192+2020,baseline,applied_cryptography,11276,0.1642,0.0868,0.0438,0.1563,0.0468,0.4327,2.971
193+2020,baseline,autonomous_scientific_discovery,193,0.1088,0.3212,0.0933,0.1813,0.1606,0.6684,4.425
194+2020,baseline,blockchain,9688,0.1226,0.0824,0.0297,0.1361,0.0325,0.4047,2.794
195+2020,post,blockchain,3,0.0,0.0,0.0,0.0,1.0,1.0,4.333
196+2020,baseline,cloud_computing,8184,0.1477,0.0842,0.0442,0.1299,0.0208,0.3877,2.905
197+2020,baseline,compilers,737,0.0909,0.19,0.0529,0.1357,0.057,0.5427,3.033
198+2020,baseline,computational_biology,18078,0.1859,0.2428,0.0534,0.2199,0.0296,0.6409,5.128
199+2020,post,computational_complexity,6,0.0,0.3333,0.0,0.1667,1.0,1.0,3.167
200+2020,pre,computational_complexity,1,0.0,0.0,0.0,1.0,1.0,1.0,4.0
201+2020,baseline,computational_complexity,3865,0.1056,0.1415,0.0168,0.1946,0.1213,0.5679,2.566
202+2020,baseline,computer_architecture,17491,0.1377,0.1962,0.0932,0.1596,0.0414,0.3303,3.686
203+2020,post,computer_graphics,1,0.0,0.0,0.0,0.0,1.0,1.0,4.0
204+2020,baseline,computer_graphics,9444,0.1364,0.144,0.0497,0.142,0.0684,0.4842,2.916
205+2020,baseline,computer_networks,54382,0.2429,0.1209,0.0662,0.2249,0.0764,0.3882,3.482
206+2020,post,computer_networks,3,0.0,0.0,0.0,0.3333,1.0,1.0,3.0
207+2020,pre,computer_networks,4,0.0,0.25,0.25,0.5,1.0,1.0,3.0
208+2020,pre,computer_vision,4,0.5,0.0,0.0,0.25,1.0,1.0,3.75
209+2020,baseline,computer_vision,84014,0.2919,0.1198,0.0653,0.1652,0.1432,0.4669,3.822
210+2020,post,computer_vision,28,0.2857,0.25,0.1786,0.2143,1.0,1.0,4.464
211+2020,baseline,cybersecurity,32697,0.1191,0.13,0.0355,0.1253,0.0651,0.4388,2.897
212+2020,post,cybersecurity,8,0.0,0.375,0.125,0.375,1.0,1.0,4.375
213+2020,post,database_systems,2,0.0,0.0,0.0,0.0,1.0,1.0,10.0
214+2020,baseline,database_systems,15057,0.1216,0.1113,0.0414,0.1343,0.0434,0.495,3.057
215+2020,pre,database_systems,2,0.0,1.0,0.0,0.5,1.0,1.0,3.0
216+2020,pre,deep_learning_architectures,2,0.0,0.0,0.0,0.0,1.0,1.0,1.5
217+2020,baseline,deep_learning_architectures,17902,0.2731,0.1615,0.0678,0.1913,0.1986,0.5126,3.727
218+2020,post,deep_learning_architectures,5,0.0,0.4,0.2,0.4,1.0,1.0,3.0
219+2020,pre,distributed_computing,2,0.5,0.5,0.0,0.5,1.0,1.0,3.5
220+2020,post,distributed_computing,1,0.0,1.0,1.0,1.0,1.0,1.0,2.0
221+2020,baseline,distributed_computing,10898,0.194,0.1899,0.0584,0.1823,0.1012,0.4681,3.447
222+2020,post,edge_computing,1,0.0,0.0,0.0,0.0,1.0,1.0,3.0
223+2020,baseline,edge_computing,7114,0.1751,0.107,0.0447,0.211,0.0283,0.3663,3.436
224+2020,baseline,federated_learning,304,0.2072,0.2072,0.0757,0.2434,0.1743,0.5461,3.905
225+2020,pre,formal_verification,2,0.5,0.0,0.0,0.5,1.0,1.0,4.5
226+2020,baseline,formal_verification,4530,0.1044,0.1415,0.0391,0.243,0.2243,0.6351,2.859
227+2020,post,formal_verification,3,0.0,0.0,0.0,0.6667,1.0,1.0,2.333
228+2020,baseline,foundation_models,137,0.3139,0.219,0.073,0.146,0.2993,0.5255,3.803
229+2020,baseline,generative_ai,1498,0.245,0.1889,0.0794,0.1462,0.2036,0.4653,3.633
230+2020,baseline,hci,36247,0.063,0.1427,0.037,0.1265,0.0322,0.4348,3.126
231+2020,post,hci,4,0.0,0.25,0.0,0.25,1.0,1.0,2.0
232+2020,post,information_retrieval,3,0.0,0.0,0.0,0.0,1.0,1.0,4.333
233+2020,baseline,information_retrieval,11774,0.214,0.1395,0.0854,0.1618,0.1181,0.4676,3.247
234+2020,baseline,llm,2461,0.1451,0.1121,0.0813,0.1077,0.0947,0.4275,3.242
235+2020,baseline,ml_theory,19292,0.1493,0.2407,0.0809,0.1813,0.3396,0.6305,3.221
236+2020,pre,ml_theory,3,0.0,0.0,0.3333,0.6667,1.0,1.0,3.667
237+2020,post,ml_theory,14,0.0,0.2143,0.0,0.1429,1.0,1.0,3.214
238+2020,baseline,multimodal_ai,755,0.4053,0.2079,0.1311,0.2291,0.1841,0.4278,4.013
239+2020,post,nlp,23,0.1739,0.1739,0.2609,0.3043,1.0,1.0,4.13
240+2020,baseline,nlp,28724,0.1429,0.1459,0.0794,0.1374,0.1809,0.5686,3.294
241+2020,baseline,operating_systems,1454,0.1534,0.1589,0.0722,0.1437,0.0303,0.3982,3.071
242+2020,post,privacy_preserving_computation,2,0.0,0.0,0.5,0.5,1.0,1.0,2.0
243+2020,baseline,privacy_preserving_computation,4155,0.1641,0.2113,0.0602,0.194,0.1232,0.5083,3.463
244+2020,post,programming_languages,3,0.0,0.0,0.0,0.3333,1.0,1.0,4.667
245+2020,baseline,programming_languages,4123,0.0589,0.155,0.0303,0.146,0.1179,0.5511,2.439
246+2020,pre,quantum_computing,3,0.0,0.6667,0.6667,0.6667,1.0,1.0,2.0
247+2020,baseline,quantum_computing,8096,0.1902,0.1784,0.0615,0.2134,0.0725,0.5983,3.697
248+2020,baseline,reinforcement_learning,5674,0.1882,0.2652,0.1038,0.1768,0.2978,0.5583,3.438
249+2020,post,reinforcement_learning,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
250+2020,post,robotics,2,0.0,0.5,0.0,0.5,1.0,1.0,2.0
251+2020,pre,robotics,5,0.2,0.8,0.4,0.4,1.0,0.8,4.0
252+2020,baseline,robotics,20346,0.1966,0.2101,0.0598,0.1794,0.1495,0.4674,3.798
253+2020,baseline,scientific_computing,19719,0.2425,0.1752,0.0312,0.2442,0.1673,0.5889,3.243
254+2020,post,scientific_computing,5,0.2,0.4,0.0,0.8,1.0,1.0,2.8
255+2020,pre,scientific_computing,16,0.0,0.125,0.0,0.375,1.0,0.9375,2.188
256+2020,post,software_engineering,1,0.0,1.0,0.0,0.0,1.0,1.0,3.0
257+2020,pre,software_engineering,1,0.0,0.0,1.0,1.0,1.0,1.0,3.0
258+2020,baseline,software_engineering,27754,0.0673,0.0862,0.0382,0.1131,0.0464,0.4258,2.662
259+2020,baseline,speech_audio,12490,0.1448,0.1029,0.0749,0.1233,0.1324,0.4532,3.205
260+2020,post,speech_audio,3,0.0,0.3333,0.3333,0.3333,1.0,1.0,2.667
261+2020,baseline,sustainable_computing,634,0.2145,0.1562,0.0568,0.2618,0.0505,0.4811,3.554
262+2021,baseline,ai_accelerators,1249,0.2858,0.2314,0.0769,0.2346,0.1153,0.4139,4.869
263+2021,post,ai_agents,2,0.0,0.5,0.0,0.5,1.0,1.0,3.0
264+2021,baseline,ai_agents,4890,0.1945,0.1415,0.0315,0.192,0.1247,0.5284,3.1
265+2021,baseline,ai_safety,2493,0.0461,0.2944,0.0542,0.1777,0.2166,0.6193,2.908
266+2021,pre,ai_safety,1,0.0,1.0,1.0,1.0,1.0,1.0,6.0
267+2021,baseline,algorithms,28971,0.2026,0.1532,0.0315,0.237,0.143,0.5746,2.96
268+2021,pre,algorithms,5,0.0,0.4,0.0,0.2,0.6,1.0,2.0
269+2021,post,algorithms,47,0.0,0.3617,0.0426,0.3191,0.9362,0.9362,3.106
270+2021,baseline,applied_cryptography,10920,0.1968,0.0904,0.0473,0.1662,0.0542,0.4442,3.1
271+2021,post,applied_cryptography,4,0.0,0.25,0.0,0.25,0.5,1.0,3.25
272+2021,baseline,autonomous_scientific_discovery,221,0.0588,0.3258,0.0814,0.1674,0.2443,0.7104,4.643
273+2021,baseline,blockchain,10644,0.1487,0.0821,0.0336,0.1458,0.0316,0.4334,2.946
274+2021,post,blockchain,3,0.3333,0.3333,0.0,0.3333,0.6667,1.0,5.333
275+2021,baseline,cloud_computing,7735,0.1694,0.0829,0.0512,0.1396,0.0244,0.4105,2.996
276+2021,baseline,compilers,709,0.1086,0.2116,0.0917,0.1636,0.0465,0.5416,3.169
277+2021,baseline,computational_biology,19218,0.2026,0.2293,0.0558,0.2368,0.0312,0.6674,5.123
278+2021,baseline,computational_complexity,3753,0.1159,0.1367,0.0189,0.1937,0.121,0.5838,2.678
279+2021,post,computational_complexity,10,0.0,0.3,0.0,0.0,0.9,0.9,2.9
280+2021,pre,computational_complexity,2,0.0,0.5,0.0,0.5,1.0,1.0,3.0
281+2021,post,computer_architecture,4,0.0,0.5,0.0,0.5,1.0,1.0,4.75
282+2021,baseline,computer_architecture,16396,0.156,0.1994,0.1045,0.1649,0.0504,0.3453,3.86
283+2021,baseline,computer_graphics,8966,0.1694,0.137,0.0552,0.1374,0.0751,0.5152,3.031
284+2021,post,computer_graphics,3,0.0,0.0,0.0,0.3333,0.6667,0.6667,2.667
285+2021,baseline,computer_networks,51635,0.2651,0.1166,0.0781,0.2306,0.083,0.4144,3.575
286+2021,pre,computer_networks,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
287+2021,post,computer_networks,8,0.125,0.0,0.125,0.25,1.0,1.0,4.125
288+2021,post,computer_vision,53,0.1887,0.1132,0.0377,0.1887,0.9245,0.9434,4.151
289+2021,baseline,computer_vision,93042,0.3358,0.1179,0.0736,0.1712,0.1556,0.4797,4.015
290+2021,pre,computer_vision,1,0.0,1.0,0.0,1.0,1.0,1.0,3.0
291+2021,baseline,cybersecurity,31770,0.138,0.1298,0.0413,0.1408,0.0716,0.4731,3.092
292+2021,post,cybersecurity,9,0.0,0.0,0.1111,0.1111,1.0,1.0,4.333
293+2021,baseline,database_systems,14331,0.1502,0.1043,0.0518,0.1423,0.0498,0.5393,3.296
294+2021,post,database_systems,1,0.0,0.0,0.0,1.0,1.0,1.0,5.0
295+2021,post,deep_learning_architectures,12,0.0833,0.1667,0.1667,0.1667,1.0,1.0,3.75
296+2021,pre,deep_learning_architectures,1,1.0,0.0,0.0,0.0,1.0,0.0,3.0
297+2021,baseline,deep_learning_architectures,21211,0.3058,0.1691,0.0764,0.2017,0.2191,0.5294,3.794
298+2021,baseline,distributed_computing,10088,0.2127,0.191,0.0643,0.1884,0.1107,0.4908,3.54
299+2021,pre,distributed_computing,1,0.0,1.0,0.0,0.0,0.0,1.0,3.0
300+2021,post,distributed_computing,7,0.1429,0.2857,0.0,0.5714,1.0,1.0,3.143
301+2021,baseline,edge_computing,7910,0.1958,0.0952,0.0497,0.2155,0.0279,0.4023,3.519
302+2021,post,federated_learning,1,0.0,0.0,0.0,0.0,1.0,1.0,4.0
303+2021,baseline,federated_learning,650,0.2677,0.2554,0.1092,0.2215,0.2385,0.5092,3.963
304+2021,post,formal_verification,4,0.0,0.0,0.0,0.5,1.0,1.0,2.5
305+2021,baseline,formal_verification,4507,0.1041,0.1378,0.0375,0.2443,0.2403,0.6357,2.842
306+2021,baseline,foundation_models,249,0.3775,0.241,0.1084,0.2209,0.3293,0.5301,4.418
307+2021,baseline,generative_ai,1959,0.2838,0.1868,0.0812,0.1741,0.2006,0.4609,3.646
308+2021,post,generative_ai,1,0.0,0.0,0.0,1.0,1.0,1.0,4.0
309+2021,post,hci,10,0.1,0.0,0.0,0.5,1.0,1.0,3.9
310+2021,baseline,hci,34419,0.0893,0.151,0.0371,0.1405,0.039,0.4914,3.338
311+2021,baseline,information_retrieval,11536,0.2499,0.1328,0.0911,0.1738,0.1351,0.4956,3.407
312+2021,post,information_retrieval,7,0.1429,0.0,0.2857,0.0,1.0,1.0,2.714
313+2021,post,llm,2,0.5,0.0,0.5,0.0,1.0,1.0,5.0
314+2021,baseline,llm,2847,0.209,0.1507,0.1117,0.1236,0.1324,0.4939,3.597
315+2021,pre,ml_theory,2,0.0,0.5,0.0,1.0,1.0,1.0,2.0
316+2021,baseline,ml_theory,19943,0.1689,0.2307,0.0824,0.1926,0.3116,0.6341,3.358
317+2021,post,ml_theory,11,0.1818,0.2727,0.0909,0.2727,1.0,1.0,3.091
318+2021,baseline,multimodal_ai,946,0.4693,0.1744,0.1332,0.2452,0.185,0.4313,4.196
319+2021,post,multimodal_ai,2,0.0,0.5,0.5,0.0,1.0,1.0,1.5
320+2021,baseline,nlp,29555,0.1756,0.1559,0.0914,0.1505,0.1871,0.6122,3.443
321+2021,post,nlp,38,0.1053,0.2632,0.1579,0.2105,0.9474,1.0,4.289
322+2021,post,operating_systems,1,0.0,0.0,0.0,0.0,0.0,0.0,1.0
323+2021,baseline,operating_systems,1322,0.18,0.1445,0.0983,0.1377,0.0272,0.4584,3.332
324+2021,post,privacy_preserving_computation,1,0.0,0.0,0.0,0.0,1.0,1.0,4.0
325+2021,baseline,privacy_preserving_computation,4240,0.1995,0.216,0.0726,0.2078,0.1363,0.5389,3.705
326+2021,baseline,programming_languages,3972,0.0622,0.1433,0.0305,0.1568,0.1365,0.5997,2.489
327+2021,post,programming_languages,6,0.0,0.0,0.0,0.1667,1.0,1.0,1.5
328+2021,baseline,quantum_computing,9273,0.1785,0.1857,0.0771,0.2061,0.0764,0.626,3.997
329+2021,pre,quantum_computing,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
330+2021,post,quantum_computing,6,0.0,0.3333,0.0,0.1667,1.0,1.0,2.333
331+2021,baseline,reinforcement_learning,6587,0.2408,0.269,0.1084,0.1861,0.3095,0.5737,3.6
332+2021,post,reinforcement_learning,5,0.0,0.0,0.2,0.0,0.8,1.0,3.0
333+2021,baseline,robotics,22059,0.2466,0.1899,0.0596,0.1815,0.1688,0.4756,3.924
334+2021,post,robotics,7,0.0,0.2857,0.0,0.0,1.0,1.0,4.143
335+2021,pre,scientific_computing,2,0.0,0.5,0.0,0.5,1.0,1.0,2.5
336+2021,post,scientific_computing,8,0.0,0.0,0.0,0.25,0.875,1.0,2.375
337+2021,baseline,scientific_computing,20289,0.2473,0.171,0.0344,0.2401,0.1854,0.5948,3.247
338+2021,post,software_engineering,7,0.1429,0.2857,0.1429,0.2857,0.8571,0.8571,4.143
339+2021,baseline,software_engineering,24664,0.0827,0.0936,0.047,0.1295,0.0722,0.4869,2.894
340+2021,baseline,speech_audio,12783,0.1848,0.1181,0.0952,0.1327,0.1519,0.476,3.468
341+2021,post,speech_audio,4,0.5,0.25,0.0,0.5,1.0,1.0,4.75
342+2021,baseline,sustainable_computing,717,0.2218,0.1855,0.0502,0.2427,0.0516,0.5202,3.661
343+2022,baseline,ai_accelerators,1341,0.2677,0.2424,0.0895,0.2528,0.1111,0.4534,5.238
344+2022,post,ai_accelerators,80,0.4875,0.225,0.0625,0.225,0.0125,0.4375,4.888
345+2022,baseline,ai_agents,4131,0.252,0.1268,0.0315,0.1915,0.1365,0.5556,3.269
346+2022,post,ai_agents,390,0.2436,0.0821,0.0308,0.1513,0.1385,0.6051,3.072
347+2022,post,ai_safety,205,0.0634,0.2585,0.0293,0.1561,0.2,0.6146,3.127
348+2022,baseline,ai_safety,2749,0.0531,0.2699,0.0691,0.1848,0.1961,0.6501,3.057
349+2022,post,algorithms,2237,0.2713,0.1203,0.0246,0.2083,0.0997,0.5431,3.077
350+2022,pre,algorithms,2,0.0,0.0,0.0,0.5,1.0,1.0,2.0
351+2022,baseline,algorithms,25963,0.2425,0.1585,0.0349,0.2437,0.1415,0.583,3.061
352+2022,baseline,applied_cryptography,9409,0.2389,0.0983,0.0498,0.186,0.0543,0.4911,3.237
353+2022,post,applied_cryptography,890,0.2742,0.0539,0.0461,0.1371,0.0371,0.5146,3.217
354+2022,baseline,autonomous_scientific_discovery,215,0.1256,0.3767,0.093,0.2093,0.1953,0.6512,4.126
355+2022,post,autonomous_scientific_discovery,21,0.0952,0.381,0.0476,0.1905,0.2857,0.7143,3.905
356+2022,post,blockchain,1169,0.1814,0.071,0.0359,0.1454,0.0257,0.4953,3.163
357+2022,baseline,blockchain,10601,0.1663,0.0874,0.0326,0.1647,0.0322,0.4817,3.052
358+2022,baseline,cloud_computing,6306,0.1917,0.0864,0.0584,0.143,0.0251,0.4565,3.124
359+2022,post,cloud_computing,797,0.1982,0.079,0.0615,0.1129,0.0163,0.4291,3.031
360+2022,post,compilers,58,0.0517,0.0345,0.0172,0.1207,0.0,0.7069,3.397
361+2022,baseline,compilers,584,0.0993,0.2072,0.0822,0.1781,0.0599,0.5976,3.433
362+2022,post,computational_biology,2018,0.2914,0.2279,0.052,0.225,0.0238,0.5956,5.172
363+2022,baseline,computational_biology,17518,0.2238,0.2211,0.0528,0.2374,0.0347,0.6935,5.352
364+2022,baseline,computational_complexity,3266,0.131,0.1562,0.0233,0.214,0.1191,0.5842,2.755
365+2022,post,computational_complexity,275,0.1309,0.1127,0.0145,0.1673,0.1382,0.6436,2.665
366+2022,post,computer_architecture,1089,0.214,0.1579,0.0817,0.1524,0.045,0.3636,3.834
367+2022,baseline,computer_architecture,14085,0.1847,0.206,0.1199,0.1693,0.0552,0.3663,4.099
368+2022,post,computer_graphics,778,0.1799,0.1015,0.063,0.1054,0.0758,0.5938,3.085
369+2022,baseline,computer_graphics,7509,0.1963,0.1472,0.0587,0.1429,0.0993,0.5683,3.186
370+2022,pre,computer_networks,1,1.0,0.0,0.0,0.0,1.0,0.0,3.0
371+2022,baseline,computer_networks,44215,0.2916,0.1113,0.0822,0.2366,0.0894,0.4315,3.75
372+2022,post,computer_networks,4925,0.317,0.0993,0.0828,0.2337,0.0404,0.3789,3.787
373+2022,post,computer_vision,8585,0.3726,0.0874,0.0515,0.1449,0.1123,0.4953,4.089
374+2022,baseline,computer_vision,90119,0.3793,0.1063,0.0745,0.1753,0.1554,0.5056,4.18
375+2022,post,cybersecurity,3238,0.1868,0.1035,0.0389,0.1242,0.0537,0.491,3.283
376+2022,baseline,cybersecurity,28316,0.157,0.1322,0.0463,0.1548,0.0784,0.513,3.24
377+2022,baseline,database_systems,12221,0.1795,0.0996,0.0493,0.15,0.0489,0.5964,3.379
378+2022,post,database_systems,1404,0.1538,0.0926,0.0513,0.1132,0.0264,0.589,3.162
379+2022,baseline,deep_learning_architectures,20266,0.3449,0.1523,0.0673,0.2006,0.2212,0.5569,3.934
380+2022,post,deep_learning_architectures,2004,0.3468,0.1208,0.0509,0.1477,0.1816,0.5334,3.88
381+2022,post,distributed_computing,855,0.3111,0.1193,0.0421,0.1509,0.076,0.4538,3.591
382+2022,baseline,distributed_computing,8524,0.2475,0.1763,0.0644,0.1953,0.1098,0.5016,3.634
383+2022,baseline,edge_computing,7666,0.2224,0.0892,0.0536,0.2237,0.0256,0.4126,3.72
384+2022,post,edge_computing,888,0.2252,0.0856,0.0417,0.2005,0.0158,0.3784,3.67
385+2022,baseline,federated_learning,880,0.3227,0.2193,0.1068,0.2568,0.2159,0.5227,4.658
386+2022,post,federated_learning,94,0.4149,0.1489,0.0426,0.2234,0.1489,0.4362,3.979
387+2022,baseline,formal_verification,3703,0.118,0.1431,0.0416,0.2684,0.2609,0.6603,3.012
388+2022,post,formal_verification,292,0.161,0.1096,0.0479,0.1952,0.1781,0.6164,3.065
389+2022,baseline,foundation_models,313,0.3674,0.2013,0.099,0.1661,0.4026,0.6006,4.629
390+2022,post,foundation_models,37,0.6216,0.0811,0.1622,0.2162,0.2162,0.6216,5.297
391+2022,post,generative_ai,208,0.2788,0.101,0.0529,0.1442,0.1394,0.4663,3.659
392+2022,baseline,generative_ai,1952,0.2874,0.144,0.0625,0.1773,0.1901,0.5195,3.781
393+2022,post,hci,2949,0.1061,0.0953,0.0173,0.1261,0.0264,0.6355,3.263
394+2022,baseline,hci,31852,0.0967,0.1548,0.0376,0.1461,0.0451,0.5358,3.491
395+2022,post,information_retrieval,908,0.2841,0.0837,0.0518,0.1355,0.0826,0.5143,3.331
396+2022,baseline,information_retrieval,10335,0.3016,0.1272,0.1087,0.1722,0.1533,0.5442,3.553
397+2022,baseline,llm,2943,0.2314,0.1699,0.1223,0.1359,0.208,0.5963,3.879
398+2022,post,llm,323,0.2322,0.1858,0.096,0.1331,0.2601,0.6347,3.96
399+2022,post,ml_theory,1918,0.1794,0.1741,0.0553,0.1496,0.2242,0.5782,3.395
400+2022,baseline,ml_theory,18213,0.1898,0.2007,0.0703,0.1872,0.2975,0.6562,3.479
401+2022,post,multimodal_ai,97,0.4845,0.1649,0.0928,0.2577,0.2165,0.4948,4.567
402+2022,baseline,multimodal_ai,1181,0.5258,0.1304,0.1431,0.1846,0.2134,0.4674,4.456
403+2022,pre,nlp,1,0.0,0.0,0.0,0.0,1.0,1.0,4.0
404+2022,baseline,nlp,26840,0.2025,0.151,0.0918,0.1623,0.1914,0.6484,3.672
405+2022,post,nlp,2837,0.203,0.1146,0.0663,0.1445,0.1981,0.6105,3.477
406+2022,baseline,operating_systems,965,0.2093,0.143,0.0777,0.1565,0.0321,0.513,3.456
407+2022,post,operating_systems,118,0.1356,0.1186,0.0339,0.1017,0.0254,0.5593,2.992
408+2022,baseline,privacy_preserving_computation,3940,0.2246,0.2152,0.0756,0.2211,0.1561,0.5817,3.769
409+2022,post,privacy_preserving_computation,436,0.3716,0.1491,0.0688,0.1881,0.1124,0.4472,3.706
410+2022,post,programming_languages,288,0.0903,0.0417,0.0208,0.1042,0.0972,0.7361,2.688
411+2022,baseline,programming_languages,3376,0.0705,0.1605,0.0427,0.1804,0.1348,0.6099,2.609
412+2022,post,quantum_computing,799,0.2115,0.199,0.0876,0.2278,0.0713,0.7134,3.946
413+2022,baseline,quantum_computing,8417,0.2053,0.2223,0.0985,0.2231,0.0926,0.672,3.959
414+2022,post,reinforcement_learning,666,0.3063,0.2042,0.0841,0.1682,0.2553,0.5225,3.728
415+2022,baseline,reinforcement_learning,6351,0.2903,0.2121,0.0931,0.1735,0.2907,0.5579,3.733
416+2022,baseline,robotics,21166,0.2672,0.1764,0.06,0.1801,0.1875,0.4943,4.074
417+2022,post,robotics,1937,0.4177,0.1239,0.0413,0.1791,0.1136,0.4399,4.08
418+2022,post,scientific_computing,1760,0.2767,0.1517,0.029,0.225,0.1261,0.5608,3.417
419+2022,pre,scientific_computing,2,0.5,0.0,0.0,0.0,1.0,1.0,1.0
420+2022,baseline,scientific_computing,18170,0.2804,0.1718,0.0369,0.251,0.2024,0.6222,3.534
421+2022,baseline,software_engineering,21619,0.0929,0.0994,0.055,0.1359,0.0741,0.5436,3.017
422+2022,post,software_engineering,2106,0.1187,0.0655,0.0418,0.0959,0.0465,0.5793,2.952
423+2022,post,speech_audio,1078,0.2746,0.0724,0.0529,0.1197,0.0891,0.551,3.432
424+2022,baseline,speech_audio,11620,0.2333,0.1095,0.0998,0.1434,0.1675,0.5202,3.643
425+2022,post,sustainable_computing,78,0.2308,0.141,0.0641,0.2308,0.0769,0.6538,3.59
426+2022,baseline,sustainable_computing,752,0.246,0.1755,0.0585,0.2447,0.0465,0.4973,3.914
427+2023,baseline,ai_accelerators,32,0.1875,0.4688,0.1875,0.3125,0.7188,0.8125,5.5
428+2023,post,ai_accelerators,1567,0.3101,0.2246,0.0862,0.238,0.0893,0.4416,5.294
429+2023,post,ai_agents,4912,0.2673,0.115,0.0364,0.1861,0.1423,0.5867,3.342
430+2023,baseline,ai_agents,79,0.3038,0.3671,0.1266,0.3924,0.8228,0.8987,4.114
431+2023,post,ai_safety,3960,0.0548,0.2593,0.0606,0.179,0.2225,0.7081,3.08
432+2023,baseline,ai_safety,55,0.0727,0.4727,0.0909,0.3273,0.8,0.9455,4.273
433+2023,pre,algorithms,1,0.0,0.0,0.0,0.0,1.0,1.0,1.0
434+2023,post,algorithms,27335,0.2562,0.1418,0.0332,0.2338,0.1265,0.6019,3.12
435+2023,baseline,algorithms,1195,0.0954,0.3247,0.0444,0.4209,0.6126,0.8435,3.09
436+2023,baseline,applied_cryptography,131,0.2366,0.1985,0.0916,0.3359,0.7405,0.8855,3.206
437+2023,post,applied_cryptography,11087,0.2385,0.0922,0.0515,0.174,0.0486,0.5315,3.256
438+2023,pre,applied_cryptography,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
439+2023,baseline,autonomous_scientific_discovery,12,0.0,0.5,0.0833,0.25,0.75,1.0,4.25
440+2023,post,autonomous_scientific_discovery,388,0.1005,0.3325,0.0696,0.2088,0.2655,0.75,4.809
441+2023,post,blockchain,14260,0.1501,0.0745,0.0304,0.1508,0.0262,0.5109,3.089
442+2023,baseline,blockchain,89,0.1573,0.2584,0.0674,0.4382,0.4045,0.7303,3.584
443+2023,post,cloud_computing,7616,0.1756,0.0865,0.0685,0.1292,0.0213,0.5169,3.11
444+2023,baseline,cloud_computing,27,0.2593,0.1852,0.0741,0.3333,0.3704,0.6667,4.111
445+2023,baseline,compilers,5,0.2,0.6,0.2,0.6,0.6,0.8,3.8
446+2023,post,compilers,687,0.131,0.1674,0.0684,0.1368,0.0509,0.6769,3.48
447+2023,baseline,computational_biology,175,0.1314,0.3886,0.1257,0.3657,0.4629,0.9029,5.869
448+2023,post,computational_biology,20401,0.2255,0.2021,0.051,0.2295,0.0383,0.6763,5.322
449+2023,baseline,computational_complexity,81,0.037,0.2593,0.0123,0.3333,0.6667,0.9259,2.247
450+2023,post,computational_complexity,3418,0.1451,0.1243,0.0237,0.2039,0.1211,0.6085,2.907
451+2023,baseline,computer_architecture,101,0.1584,0.4257,0.1881,0.3267,0.7426,0.8614,5.366
452+2023,post,computer_architecture,16268,0.2064,0.1934,0.1127,0.1679,0.0562,0.4078,4.182
453+2023,pre,computer_graphics,1,1.0,0.0,0.0,1.0,1.0,1.0,4.0
454+2023,post,computer_graphics,9201,0.2016,0.1387,0.0619,0.1362,0.1048,0.6405,3.25
455+2023,baseline,computer_graphics,95,0.2421,0.2947,0.1368,0.3684,0.5263,0.7684,4.126
456+2023,post,computer_networks,51205,0.305,0.1044,0.0872,0.2313,0.0719,0.4381,3.854
457+2023,baseline,computer_networks,676,0.2988,0.2396,0.1272,0.4911,0.8121,0.895,4.087
458+2023,baseline,computer_vision,2547,0.36,0.256,0.1841,0.3286,0.8041,0.883,5.142
459+2023,post,computer_vision,111714,0.3907,0.095,0.0672,0.164,0.1516,0.522,4.276
460+2023,baseline,cybersecurity,368,0.1929,0.3478,0.106,0.3723,0.7853,0.894,4.315
461+2023,post,cybersecurity,36113,0.1543,0.1243,0.0435,0.1435,0.0674,0.5534,3.26
462+2023,baseline,database_systems,95,0.1789,0.2842,0.1474,0.4526,0.8,0.9474,4.695
463+2023,post,database_systems,14392,0.1695,0.0933,0.0538,0.1336,0.0456,0.6591,3.392
464+2023,baseline,deep_learning_architectures,629,0.2814,0.3084,0.1113,0.3037,0.8124,0.9094,4.161
465+2023,post,deep_learning_architectures,24978,0.3643,0.1375,0.068,0.1931,0.1961,0.5627,3.985
466+2023,post,distributed_computing,9664,0.2739,0.1739,0.0691,0.1911,0.0972,0.5465,3.73
467+2023,baseline,distributed_computing,134,0.1567,0.3955,0.1269,0.3881,0.7239,0.8806,4.94
468+2023,post,edge_computing,9080,0.2148,0.0878,0.051,0.2214,0.0246,0.43,3.721
469+2023,baseline,edge_computing,40,0.25,0.4,0.1,0.475,0.7,0.825,5.125
470+2023,post,federated_learning,1467,0.3565,0.182,0.0832,0.2522,0.1677,0.5058,4.235
471+2023,baseline,federated_learning,30,0.4667,0.2667,0.0667,0.5333,0.7667,0.8333,4.867
472+2023,baseline,formal_verification,137,0.0511,0.2774,0.0438,0.4088,0.7664,0.9489,3.051
473+2023,post,formal_verification,4248,0.1236,0.137,0.0391,0.2651,0.2651,0.6883,2.972
474+2023,post,foundation_models,563,0.3783,0.1829,0.0853,0.1741,0.444,0.6643,5.025
475+2023,baseline,foundation_models,8,0.25,0.25,0.125,0.375,0.625,0.625,4.0
476+2023,baseline,generative_ai,49,0.1429,0.3061,0.1633,0.2653,0.7143,0.7755,3.796
477+2023,post,generative_ai,3597,0.2616,0.1507,0.0556,0.1529,0.1724,0.5635,3.571
478+2023,post,hci,40423,0.0994,0.1392,0.0352,0.1424,0.0483,0.6369,3.489
479+2023,baseline,hci,164,0.1037,0.3598,0.0915,0.3476,0.5793,0.8902,4.622
480+2023,post,information_retrieval,12421,0.3156,0.1169,0.1065,0.1647,0.1479,0.6093,3.601
481+2023,baseline,information_retrieval,147,0.3673,0.2925,0.2585,0.3673,0.7959,0.9048,5.224
482+2023,post,llm,10332,0.1815,0.2002,0.0938,0.1495,0.2537,0.7129,3.78
483+2023,baseline,llm,57,0.3158,0.4737,0.2982,0.2982,0.7018,0.8947,12.579
484+2023,baseline,ml_theory,733,0.1514,0.3779,0.1091,0.3302,0.7422,0.9018,3.626
485+2023,post,ml_theory,22160,0.1991,0.1787,0.0602,0.1764,0.2568,0.6606,3.526
486+2023,baseline,multimodal_ai,18,0.4444,0.2222,0.1667,0.1111,0.7222,0.9444,5.5
487+2023,post,multimodal_ai,1914,0.5486,0.1332,0.1254,0.187,0.2665,0.5946,4.753
488+2023,post,nlp,33400,0.2043,0.1387,0.0764,0.1607,0.2012,0.6816,3.664
489+2023,baseline,nlp,565,0.2142,0.3504,0.2265,0.2991,0.6708,0.9593,5.228
490+2023,baseline,operating_systems,3,0.6667,0.0,0.3333,0.0,1.0,1.0,6.0
491+2023,post,operating_systems,1170,0.2308,0.1521,0.0897,0.141,0.0504,0.5718,3.696
492+2023,baseline,privacy_preserving_computation,104,0.1923,0.4231,0.1442,0.5096,0.7115,0.9231,4.394
493+2023,post,privacy_preserving_computation,4906,0.2662,0.1936,0.0671,0.2224,0.1449,0.5946,3.818
494+2023,post,programming_languages,3822,0.0801,0.1308,0.0322,0.1533,0.1227,0.6669,2.567
495+2023,baseline,programming_languages,54,0.037,0.3519,0.0556,0.3704,0.7037,0.9444,3.593
496+2023,post,quantum_computing,10237,0.1889,0.2151,0.0962,0.2131,0.0795,0.6837,4.082
497+2023,baseline,quantum_computing,307,0.1629,0.316,0.1531,0.3974,0.4235,0.9283,4.798
498+2023,post,reinforcement_learning,8203,0.3344,0.1872,0.0745,0.1681,0.2604,0.5574,3.832
499+2023,baseline,reinforcement_learning,219,0.1644,0.3744,0.137,0.2877,0.79,0.8767,3.699
500+2023,baseline,robotics,505,0.1505,0.4376,0.1168,0.2653,0.8733,0.9267,4.364
501+2023,post,robotics,25254,0.3114,0.1607,0.0569,0.1764,0.1845,0.5093,4.152
502+2023,post,scientific_computing,19480,0.2953,0.1676,0.0401,0.2462,0.1657,0.6018,3.429
503+2023,pre,scientific_computing,3,0.0,0.0,0.0,0.6667,1.0,1.0,2.0
504+2023,baseline,scientific_computing,940,0.1628,0.334,0.0245,0.3543,0.834,0.9489,3.152
505+2023,baseline,software_engineering,187,0.1711,0.1872,0.1337,0.4225,0.5348,0.8235,4.294
506+2023,post,software_engineering,26058,0.0995,0.0924,0.0499,0.1292,0.0722,0.6213,3.019
507+2023,baseline,speech_audio,355,0.1972,0.3577,0.338,0.2958,0.8254,0.8958,4.766
508+2023,post,speech_audio,14333,0.2172,0.0996,0.0837,0.1322,0.1704,0.5598,3.626
509+2023,post,sustainable_computing,1073,0.2432,0.1622,0.0801,0.2209,0.0718,0.5778,3.954
510+2023,baseline,sustainable_computing,5,0.0,0.6,0.2,0.0,1.0,1.0,3.8
511+2024,baseline,ai_accelerators,5,0.0,0.2,0.2,0.4,1.0,1.0,5.8
512+2024,post,ai_accelerators,1715,0.3376,0.193,0.0746,0.2117,0.1073,0.4239,5.27
513+2024,post,ai_agents,5888,0.2661,0.1165,0.0343,0.1619,0.1686,0.5922,3.481
514+2024,baseline,ai_agents,11,0.3636,0.1818,0.0909,0.2727,0.9091,1.0,4.091
515+2024,baseline,ai_safety,16,0.1875,0.5625,0.125,0.25,0.625,0.75,3.75
516+2024,post,ai_safety,6201,0.0555,0.2348,0.0437,0.168,0.2092,0.7002,3.166
517+2024,baseline,algorithms,301,0.1063,0.3023,0.0399,0.4219,0.6113,0.8937,3.086
518+2024,pre,algorithms,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
519+2024,post,algorithms,29655,0.2628,0.1426,0.0331,0.2325,0.1478,0.6008,3.178
520+2024,baseline,applied_cryptography,18,0.3333,0.2778,0.0,0.5556,0.6111,0.9444,3.889
521+2024,post,applied_cryptography,12408,0.2445,0.0892,0.0497,0.172,0.0516,0.4971,3.358
522+2024,post,autonomous_scientific_discovery,456,0.1316,0.2368,0.0658,0.2061,0.25,0.7193,4.524
523+2024,baseline,blockchain,13,0.1538,0.3077,0.0,0.3846,0.3846,0.8462,4.0
524+2024,post,blockchain,15363,0.1381,0.0738,0.0235,0.142,0.0258,0.5146,3.143
525+2024,post,cloud_computing,8597,0.1599,0.0976,0.0776,0.128,0.0264,0.5386,3.166
526+2024,baseline,cloud_computing,12,0.1667,0.5,0.0,0.3333,0.4167,0.6667,4.0
527+2024,baseline,compilers,1,0.0,0.0,0.0,0.0,1.0,1.0,7.0
528+2024,post,compilers,767,0.146,0.1669,0.09,0.1669,0.0808,0.6454,3.568
529+2024,post,computational_biology,23499,0.2475,0.1911,0.0466,0.2245,0.0485,0.6748,5.41
530+2024,baseline,computational_biology,23,0.1739,0.4348,0.2609,0.6087,0.5217,0.9565,5.696
531+2024,baseline,computational_complexity,37,0.0811,0.2432,0.027,0.3243,0.5405,0.8108,2.757
532+2024,post,computational_complexity,3787,0.1627,0.122,0.019,0.1954,0.1336,0.6237,2.962
533+2024,post,computer_architecture,17149,0.2345,0.1865,0.1051,0.1558,0.0735,0.3984,4.488
534+2024,baseline,computer_architecture,10,0.1,0.4,0.0,0.4,0.8,1.0,7.7
535+2024,baseline,computer_graphics,21,0.1905,0.1429,0.1429,0.3333,0.5238,0.7619,4.19
536+2024,post,computer_graphics,10590,0.216,0.128,0.0579,0.1317,0.1256,0.6182,3.385
537+2024,baseline,computer_networks,103,0.2621,0.2136,0.1456,0.4757,0.6505,0.7961,4.049
538+2024,post,computer_networks,54239,0.3051,0.1057,0.0762,0.2299,0.0843,0.4255,3.934
539+2024,baseline,computer_vision,375,0.4613,0.2533,0.1573,0.3787,0.7173,0.84,5.067
540+2024,post,computer_vision,132942,0.4018,0.0919,0.0593,0.1611,0.1741,0.52,4.4
541+2024,baseline,cybersecurity,69,0.3043,0.3188,0.1159,0.3913,0.6667,0.8841,4.812
542+2024,post,cybersecurity,45479,0.1501,0.1174,0.0405,0.1389,0.0753,0.5558,3.311
543+2024,post,database_systems,15014,0.1659,0.097,0.0515,0.1251,0.0565,0.6648,3.497
544+2024,baseline,database_systems,10,0.2,0.3,0.0,0.3,0.7,0.9,3.9
545+2024,post,deep_learning_architectures,28853,0.3714,0.1316,0.0589,0.1904,0.2177,0.5622,4.154
546+2024,baseline,deep_learning_architectures,132,0.3182,0.303,0.0909,0.3258,0.7576,0.9015,4.129
547+2024,baseline,distributed_computing,27,0.1481,0.2222,0.1111,0.3333,0.5556,0.8148,3.963
548+2024,post,distributed_computing,10574,0.2745,0.1701,0.0681,0.1838,0.1154,0.526,3.864
549+2024,baseline,edge_computing,7,0.1429,0.5714,0.2857,0.4286,0.5714,0.7143,4.286
550+2024,post,edge_computing,10203,0.2225,0.0828,0.0465,0.2039,0.0297,0.4136,3.778
551+2024,post,federated_learning,2065,0.4232,0.1554,0.0634,0.2615,0.1758,0.4552,4.311
552+2024,baseline,federated_learning,5,0.0,0.4,0.4,0.2,0.6,0.8,3.8
553+2024,pre,formal_verification,1,0.0,0.0,0.0,0.0,1.0,0.0,1.0
554+2024,post,formal_verification,4194,0.1319,0.1328,0.042,0.2737,0.253,0.695,3.069
555+2024,baseline,formal_verification,38,0.0263,0.1842,0.0263,0.4737,0.8421,0.9737,2.842
556+2024,baseline,foundation_models,1,1.0,1.0,0.0,1.0,1.0,1.0,4.0
557+2024,post,foundation_models,1045,0.3407,0.2364,0.0947,0.2335,0.4871,0.6861,5.201
558+2024,baseline,generative_ai,11,0.2727,0.0909,0.0909,0.3636,0.8182,1.0,3.636
559+2024,post,generative_ai,7136,0.2379,0.1359,0.0425,0.1382,0.1392,0.5646,3.873
560+2024,baseline,hci,32,0.1875,0.2813,0.125,0.5625,0.4688,0.8438,4.906
561+2024,post,hci,48042,0.1042,0.1352,0.0298,0.1373,0.0617,0.6278,3.52
562+2024,post,information_retrieval,14120,0.3161,0.1125,0.0934,0.1618,0.1756,0.5898,3.642
563+2024,baseline,information_retrieval,26,0.3846,0.2308,0.2308,0.3846,0.8077,0.8462,4.577
564+2024,post,llm,24861,0.173,0.1976,0.0856,0.1443,0.2912,0.7103,4.11
565+2024,baseline,llm,12,0.3333,0.3333,0.25,0.4167,0.3333,0.6667,5.167
566+2024,baseline,ml_theory,179,0.162,0.3911,0.067,0.3855,0.6983,0.8994,3.626
567+2024,post,ml_theory,26531,0.1971,0.1731,0.0525,0.1778,0.2528,0.6623,3.573
568+2024,post,multimodal_ai,3402,0.525,0.137,0.0964,0.1999,0.3136,0.5632,5.121
569+2024,baseline,multimodal_ai,6,1.0,0.1667,0.1667,0.6667,0.6667,0.8333,5.833
570+2024,baseline,nlp,74,0.3108,0.2703,0.2027,0.2568,0.5946,0.8514,5.122
571+2024,post,nlp,38757,0.1934,0.1256,0.0565,0.1473,0.208,0.6729,3.796
572+2024,post,operating_systems,1259,0.2367,0.1374,0.0874,0.1398,0.0445,0.5409,3.742
573+2024,baseline,operating_systems,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
574+2024,baseline,privacy_preserving_computation,18,0.1111,0.6111,0.2778,0.3889,0.6111,1.0,4.444
575+2024,post,privacy_preserving_computation,6668,0.2722,0.1752,0.0604,0.2037,0.1435,0.5775,3.819
576+2024,baseline,programming_languages,8,0.0,0.125,0.125,0.375,0.75,0.75,4.25
577+2024,post,programming_languages,3838,0.08,0.149,0.0323,0.1628,0.1397,0.6709,2.706
578+2024,baseline,quantum_computing,75,0.1467,0.48,0.1733,0.4267,0.4133,0.92,4.44
579+2024,post,quantum_computing,12647,0.1791,0.215,0.0924,0.2163,0.0991,0.6726,4.147
580+2024,baseline,reinforcement_learning,46,0.1739,0.4348,0.1304,0.3043,0.7609,0.8261,3.957
581+2024,post,reinforcement_learning,9846,0.3523,0.1701,0.0639,0.172,0.2556,0.5581,3.828
582+2024,post,robotics,29916,0.3173,0.1614,0.0531,0.1716,0.2201,0.5058,4.302
583+2024,baseline,robotics,70,0.1429,0.4,0.1,0.2714,0.8286,0.9286,4.586
584+2024,baseline,scientific_computing,161,0.2609,0.3168,0.0435,0.3975,0.7081,0.8882,3.081
585+2024,post,scientific_computing,21726,0.3014,0.1647,0.0369,0.2337,0.2144,0.6183,3.49
586+2024,post,software_engineering,30489,0.1064,0.0944,0.0501,0.1222,0.0907,0.636,3.183
587+2024,baseline,software_engineering,44,0.0682,0.1591,0.1136,0.4318,0.5455,0.7727,4.159
588+2024,baseline,speech_audio,23,0.2174,0.1739,0.087,0.3478,0.4783,0.7826,4.435
589+2024,post,speech_audio,16404,0.2245,0.0943,0.0739,0.1312,0.1822,0.5574,3.721
590+2024,post,sustainable_computing,1271,0.2557,0.1338,0.0795,0.1943,0.0787,0.5578,3.889
591+2025,post,ai_accelerators,1906,0.351,0.1406,0.0714,0.1768,0.1081,0.4481,5.341
592+2025,post,ai_agents,8815,0.2288,0.1105,0.04,0.1416,0.2358,0.6535,3.683
593+2025,baseline,ai_agents,4,0.0,0.75,0.0,0.5,1.0,1.0,2.5
594+2025,post,ai_safety,10881,0.0543,0.1753,0.0297,0.1442,0.202,0.709,3.128
595+2025,baseline,ai_safety,4,0.0,0.5,0.25,0.5,1.0,1.0,2.25
596+2025,post,algorithms,33709,0.252,0.1255,0.0281,0.2162,0.1636,0.6326,3.179
597+2025,baseline,algorithms,86,0.093,0.2442,0.0233,0.4186,0.8488,1.0,2.965
598+2025,baseline,applied_cryptography,5,0.2,0.2,0.0,0.2,1.0,1.0,2.4
599+2025,post,applied_cryptography,14607,0.2125,0.0813,0.0509,0.1545,0.0637,0.5226,3.311
600+2025,post,autonomous_scientific_discovery,764,0.1414,0.1872,0.0537,0.1767,0.3141,0.7853,5.212
601+2025,post,blockchain,18011,0.1128,0.0705,0.0241,0.1341,0.0253,0.5348,3.069
602+2025,baseline,blockchain,1,0.0,0.0,0.0,1.0,1.0,1.0,3.0
603+2025,post,cloud_computing,10973,0.1211,0.1111,0.0722,0.1008,0.0327,0.5728,2.922
604+2025,baseline,cloud_computing,1,1.0,0.0,0.0,1.0,1.0,1.0,4.0
605+2025,post,compilers,767,0.1408,0.1499,0.0508,0.1499,0.0782,0.6597,3.58
606+2025,baseline,computational_biology,2,0.5,0.5,0.0,0.0,1.0,1.0,3.0
607+2025,post,computational_biology,28152,0.2449,0.1696,0.0385,0.2155,0.0575,0.6719,5.511
608+2025,baseline,computational_complexity,12,0.0833,0.25,0.0833,0.4167,1.0,1.0,2.667
609+2025,post,computational_complexity,4197,0.1384,0.1044,0.0157,0.1978,0.1592,0.6652,2.894
610+2025,baseline,computer_architecture,1,0.0,1.0,0.0,0.0,1.0,1.0,6.0
611+2025,post,computer_architecture,20208,0.2336,0.1645,0.0941,0.1425,0.0947,0.4181,4.374
612+2025,post,computer_graphics,12439,0.2169,0.1149,0.053,0.1274,0.1329,0.6286,3.498
613+2025,baseline,computer_graphics,1,0.0,0.0,0.0,1.0,1.0,1.0,6.0
614+2025,post,computer_networks,59546,0.3019,0.094,0.0708,0.2197,0.086,0.4309,3.999
615+2025,baseline,computer_networks,9,0.4444,0.3333,0.0,0.4444,1.0,1.0,3.333
616+2025,post,computer_vision,150701,0.384,0.0821,0.0528,0.1519,0.1899,0.5276,4.528
617+2025,baseline,computer_vision,56,0.3929,0.3214,0.1964,0.4643,1.0,1.0,4.804
618+2025,post,cybersecurity,59026,0.1374,0.1044,0.0351,0.1254,0.0853,0.5625,3.281
619+2025,baseline,cybersecurity,18,0.2778,0.2778,0.0556,0.3889,1.0,1.0,3.778
620+2025,baseline,database_systems,7,0.2857,0.4286,0.1429,0.2857,1.0,1.0,4.571
621+2025,post,database_systems,16674,0.1328,0.0945,0.0506,0.1136,0.0678,0.7097,3.437
622+2025,post,deep_learning_architectures,34110,0.3518,0.105,0.0488,0.1734,0.2299,0.58,3.985
623+2025,baseline,deep_learning_architectures,31,0.2903,0.1935,0.0645,0.4839,0.9677,1.0,3.677
624+2025,post,distributed_computing,12541,0.2456,0.1487,0.0657,0.1604,0.1293,0.5647,4.068
625+2025,pre,distributed_computing,1,0.0,0.0,0.0,0.0,1.0,1.0,1.0
626+2025,baseline,distributed_computing,8,0.375,0.25,0.0,0.25,1.0,1.0,3.25
627+2025,post,edge_computing,11554,0.2095,0.082,0.0413,0.1895,0.0286,0.4313,3.748
628+2025,baseline,federated_learning,1,0.0,0.0,0.0,0.0,1.0,1.0,6.0
629+2025,post,federated_learning,2543,0.4019,0.1074,0.0448,0.2206,0.1479,0.44,4.244
630+2025,baseline,formal_verification,10,0.1,0.2,0.1,0.4,0.9,1.0,3.6
631+2025,post,formal_verification,4613,0.1108,0.1288,0.0334,0.253,0.3213,0.7175,3.036
632+2025,baseline,foundation_models,1,1.0,0.0,0.0,0.0,1.0,1.0,3.0
633+2025,post,foundation_models,1853,0.3092,0.1641,0.0912,0.1878,0.4873,0.6999,7.008
634+2025,baseline,generative_ai,2,0.0,0.0,0.0,0.0,1.0,1.0,2.0
635+2025,post,generative_ai,12741,0.232,0.1183,0.0352,0.1352,0.1188,0.5894,3.315
636+2025,baseline,hci,4,0.5,0.0,0.25,1.0,1.0,1.0,5.25
637+2025,post,hci,58144,0.0952,0.1144,0.0237,0.1274,0.0696,0.6531,3.478
638+2025,baseline,information_retrieval,5,0.2,0.4,0.2,0.6,1.0,1.0,3.8
639+2025,post,information_retrieval,17374,0.283,0.0894,0.0706,0.1377,0.1806,0.6215,3.698
640+2025,baseline,llm,1,0.0,1.0,0.0,0.0,1.0,1.0,3.0
641+2025,post,llm,45157,0.1716,0.1428,0.0652,0.1272,0.2818,0.7223,4.274
642+2025,post,ml_theory,32743,0.1822,0.1381,0.0401,0.1614,0.246,0.672,3.503
643+2025,baseline,ml_theory,55,0.1636,0.4727,0.0909,0.4727,0.8364,1.0,3.273
644+2025,post,multimodal_ai,6749,0.4749,0.1193,0.0794,0.1886,0.3063,0.5709,5.353
645+2025,baseline,nlp,13,0.2308,0.1538,0.1538,0.3077,0.9231,1.0,6.308
646+2025,post,nlp,42901,0.1673,0.1029,0.0449,0.1313,0.1974,0.6974,3.847
647+2025,post,operating_systems,1363,0.2201,0.1409,0.102,0.1394,0.0741,0.5708,3.864
648+2025,baseline,privacy_preserving_computation,4,0.25,0.5,0.5,0.5,1.0,1.0,3.25
649+2025,post,privacy_preserving_computation,8813,0.248,0.145,0.0478,0.1896,0.1317,0.5814,3.765
650+2025,post,programming_languages,4190,0.0766,0.1344,0.0258,0.1492,0.1618,0.6912,2.693
651+2025,baseline,programming_languages,3,0.0,0.0,0.0,0.0,0.6667,1.0,1.0
652+2025,post,quantum_computing,17457,0.1445,0.1819,0.0782,0.1878,0.1051,0.6794,4.136
653+2025,baseline,quantum_computing,14,0.1429,0.3571,0.0,0.5714,0.5,1.0,4.071
654+2025,post,reinforcement_learning,13202,0.3453,0.1331,0.0507,0.1556,0.2486,0.564,3.872
655+2025,baseline,reinforcement_learning,8,0.375,0.5,0.375,0.25,0.875,1.0,2.875
656+2025,baseline,robotics,13,0.2308,0.3846,0.1538,0.3846,1.0,1.0,4.077
657+2025,post,robotics,33451,0.334,0.1384,0.0459,0.1619,0.2615,0.5187,4.502
658+2025,baseline,scientific_computing,25,0.2,0.44,0.04,0.48,0.8,1.0,3.04
659+2025,post,scientific_computing,23163,0.277,0.1471,0.0314,0.2194,0.2629,0.6314,3.489
660+2025,pre,scientific_computing,3,0.3333,0.0,0.0,0.3333,1.0,1.0,2.667
661+2025,baseline,software_engineering,5,0.0,0.0,0.0,0.6,1.0,1.0,3.4
662+2025,post,software_engineering,36376,0.0962,0.0879,0.0416,0.1078,0.1122,0.6499,3.196
663+2025,post,speech_audio,18372,0.2127,0.0741,0.0554,0.1185,0.1869,0.5608,3.854
664+2025,baseline,speech_audio,1,0.0,1.0,0.0,0.0,1.0,1.0,4.0
665+2025,post,sustainable_computing,1639,0.1751,0.1086,0.0586,0.1733,0.0763,0.6223,3.628
666+2026,post,ai_accelerators,1649,0.336,0.174,0.0491,0.1734,0.1552,0.5221,5.324
667+2026,baseline,ai_agents,2,0.0,0.0,0.0,0.5,1.0,1.0,2.5
668+2026,post,ai_agents,12207,0.2079,0.1603,0.0646,0.156,0.3532,0.7345,3.959
669+2026,post,ai_safety,12544,0.0608,0.1738,0.0316,0.1383,0.2433,0.7451,3.152
670+2026,post,algorithms,28744,0.2104,0.1472,0.029,0.2084,0.2672,0.7478,3.014
671+2026,baseline,algorithms,32,0.0625,0.3125,0.0938,0.375,0.9375,1.0,3.438
672+2026,post,applied_cryptography,12176,0.2034,0.0786,0.042,0.1438,0.0954,0.5779,3.224
673+2026,baseline,applied_cryptography,2,0.0,0.0,0.0,0.0,1.0,1.0,2.5
674+2026,post,autonomous_scientific_discovery,910,0.178,0.2176,0.0769,0.1615,0.4846,0.8253,5.401
675+2026,post,blockchain,13276,0.0923,0.0696,0.0158,0.1114,0.0316,0.5505,2.83
676+2026,post,cloud_computing,8774,0.1039,0.0871,0.052,0.0888,0.0482,0.5391,2.806
677+2026,post,compilers,621,0.124,0.1787,0.0709,0.1417,0.1514,0.7134,3.467
678+2026,pre,computational_biology,1,0.0,0.0,0.0,1.0,1.0,1.0,2.0
679+2026,post,computational_biology,23786,0.2089,0.1665,0.0377,0.2018,0.0782,0.7036,5.376
680+2026,post,computational_complexity,4066,0.0878,0.1151,0.015,0.1626,0.2445,0.8195,2.526
681+2026,baseline,computational_complexity,1,0.0,0.0,0.0,0.0,1.0,1.0,1.0
682+2026,baseline,computer_architecture,1,0.0,1.0,0.0,1.0,1.0,1.0,4.0
683+2026,post,computer_architecture,15793,0.2125,0.1722,0.0806,0.1376,0.1615,0.5432,4.239
684+2026,post,computer_graphics,10258,0.1866,0.0979,0.0441,0.1096,0.1741,0.7021,3.48
685+2026,baseline,computer_graphics,2,0.5,0.5,0.0,0.0,1.0,1.0,5.0
686+2026,post,computer_networks,42412,0.2719,0.0893,0.0573,0.2024,0.1242,0.4981,3.849
687+2026,baseline,computer_networks,4,0.0,0.5,0.0,0.5,1.0,1.0,2.25
688+2026,post,computer_vision,117896,0.3551,0.0818,0.049,0.1483,0.2417,0.5989,4.564
689+2026,baseline,computer_vision,15,0.4,0.1333,0.0,0.2,1.0,1.0,3.867
690+2026,post,cybersecurity,51616,0.1206,0.1037,0.0311,0.1147,0.1273,0.6154,3.168
691+2026,baseline,cybersecurity,2,1.0,0.0,0.0,0.5,1.0,1.0,6.0
692+2026,baseline,database_systems,1,1.0,1.0,0.0,1.0,1.0,1.0,7.0
693+2026,post,database_systems,14094,0.1137,0.0828,0.0474,0.1003,0.0968,0.7181,3.269
694+2026,baseline,deep_learning_architectures,10,0.3,0.2,0.0,0.4,1.0,1.0,3.4
695+2026,post,deep_learning_architectures,26143,0.2987,0.1228,0.0433,0.1686,0.3059,0.6849,3.797
696+2026,baseline,distributed_computing,6,0.1667,0.3333,0.1667,1.0,1.0,1.0,6.0
697+2026,post,distributed_computing,9996,0.2136,0.1585,0.0657,0.1464,0.2187,0.6822,3.691
698+2026,post,edge_computing,8718,0.1969,0.0724,0.0349,0.1734,0.0373,0.464,3.659
699+2026,post,federated_learning,2841,0.3752,0.1059,0.0338,0.2214,0.1309,0.4368,4.139
700+2026,baseline,formal_verification,1,0.0,0.0,0.0,0.0,1.0,1.0,2.0
701+2026,pre,formal_verification,1,0.0,0.0,0.0,0.0,1.0,1.0,3.0
702+2026,post,formal_verification,4864,0.1059,0.1295,0.0315,0.2243,0.3951,0.7932,2.89
703+2026,post,foundation_models,2688,0.2913,0.1972,0.0792,0.2102,0.5547,0.7879,5.634
704+2026,post,generative_ai,14511,0.1911,0.1083,0.0256,0.1209,0.1264,0.6498,3.177
705+2026,baseline,hci,3,0.0,0.6667,0.0,1.0,1.0,1.0,5.0
706+2026,post,hci,47216,0.0924,0.1211,0.022,0.1243,0.0952,0.7226,3.415
707+2026,post,information_retrieval,16901,0.2574,0.116,0.0983,0.1348,0.2389,0.7213,3.92
708+2026,baseline,llm,2,0.0,0.0,0.0,0.0,1.0,1.0,1.0
709+2026,post,llm,51452,0.1978,0.1728,0.0803,0.1447,0.3587,0.8004,4.187
710+2026,post,ml_theory,29441,0.178,0.1632,0.0424,0.164,0.3013,0.7548,3.52
711+2026,baseline,ml_theory,16,0.1875,0.4375,0.3125,0.375,1.0,1.0,3.625
712+2026,post,multimodal_ai,8806,0.433,0.1296,0.0781,0.191,0.3541,0.6757,5.221
713+2026,post,nlp,34784,0.1522,0.1131,0.0478,0.1346,0.2455,0.7464,3.857
714+2026,baseline,nlp,1,0.0,1.0,0.0,0.0,1.0,1.0,1.0
715+2026,post,operating_systems,1247,0.2109,0.1195,0.0834,0.1171,0.1171,0.66,3.48
716+2026,post,privacy_preserving_computation,7624,0.2152,0.1516,0.0403,0.1838,0.1613,0.6381,3.716
717+2026,baseline,privacy_preserving_computation,1,1.0,0.0,0.0,0.0,1.0,1.0,4.0
718+2026,post,programming_languages,3689,0.0716,0.1342,0.0279,0.1458,0.2372,0.791,2.586
719+2026,baseline,programming_languages,2,0.5,0.0,0.0,0.5,1.0,1.0,4.5
720+2026,baseline,quantum_computing,4,0.0,0.75,0.25,0.25,0.5,1.0,4.75
721+2026,post,quantum_computing,19268,0.1454,0.1911,0.0757,0.1967,0.1269,0.7802,3.953
722+2026,post,reinforcement_learning,12551,0.3249,0.1582,0.0658,0.1591,0.3488,0.6796,3.962
723+2026,baseline,reinforcement_learning,4,0.25,0.0,0.0,0.25,1.0,1.0,3.75
724+2026,baseline,robotics,2,0.5,0.0,0.0,0.0,1.0,1.0,4.5
725+2026,post,robotics,27672,0.3049,0.1535,0.0534,0.1627,0.4274,0.6469,4.875
726+2026,baseline,scientific_computing,7,0.2857,0.4286,0.0,0.2857,0.7143,1.0,2.714
727+2026,post,scientific_computing,18731,0.2406,0.1702,0.0299,0.217,0.4174,0.7527,3.33
728+2026,post,software_engineering,34259,0.1017,0.1019,0.046,0.112,0.1823,0.7272,3.16
729+2026,baseline,software_engineering,2,0.0,0.0,0.0,0.0,1.0,1.0,3.5
730+2026,post,speech_audio,15624,0.2118,0.1025,0.0739,0.1411,0.2551,0.6826,3.905
731+2026,post,sustainable_computing,1328,0.1416,0.1152,0.0361,0.1521,0.1273,0.6905,3.414
added results/bibliometrics/category_impact_yearly.csv +354 −0
@@ -0,0 +1,354 @@
1+year,primary_category,n,c36_complete,mean_c36,med_c36,share_uncited_36,mean_ncs36,gini_c36,share_top10_fwci
2+2018,ai_accelerators,816,true,8.289,2.0,0.2745,1.327,0.757,0.212
3+2018,ai_agents,5433,true,3.72,1.0,0.4813,0.765,0.7954,0.1178
4+2018,ai_safety,1052,true,8.545,1.0,0.4163,1.323,0.8437,0.3061
5+2018,algorithms,29121,true,3.818,1.0,0.4019,0.842,0.7527,0.1296
6+2018,applied_cryptography,10762,true,3.689,1.0,0.4835,0.783,0.8065,0.093
7+2018,blockchain,6434,true,5.917,0.0,0.5508,2.663,0.8687,0.1993
8+2018,cloud_computing,9376,true,2.882,0.0,0.5385,1.161,0.8133,0.1525
9+2018,compilers,724,true,3.349,0.0,0.5331,0.845,0.8217,0.1215
10+2018,computational_biology,12959,true,10.831,3.0,0.2682,1.404,0.7825,0.2343
11+2018,computational_complexity,3835,true,2.534,0.0,0.5001,0.642,0.783,0.0881
12+2018,computer_architecture,19544,true,3.273,1.0,0.4311,0.879,0.7532,0.1148
13+2018,computer_graphics,10190,true,2.671,0.0,0.5963,0.94,0.8504,0.1194
14+2018,computer_networks,58645,true,5.449,1.0,0.371,1.083,0.766,0.1561
15+2018,computer_vision,69605,true,7.758,1.0,0.3809,1.226,0.8249,0.1559
16+2018,cybersecurity,29391,true,3.979,0.0,0.5146,1.169,0.8337,0.1401
17+2018,database_systems,16241,true,2.72,0.0,0.5508,0.834,0.8233,0.1204
18+2018,deep_learning_architectures,11975,true,9.209,2.0,0.3434,1.616,0.8027,0.1994
19+2018,distributed_computing,11511,true,4.144,1.0,0.4332,1.263,0.7933,0.1271
20+2018,edge_computing,5174,true,7.987,2.0,0.3307,2.129,0.7727,0.2733
21+2018,formal_verification,5307,true,3.7,1.0,0.3784,1.129,0.7463,0.1029
22+2018,generative_ai,765,true,10.68,2.0,0.3399,1.526,0.8104,0.1542
23+2018,hci,34208,true,3.264,0.0,0.5059,0.933,0.8026,0.1629
24+2018,information_retrieval,12086,true,3.979,1.0,0.4757,1.355,0.8108,0.1742
25+2018,llm,1633,true,2.124,0.0,0.5922,0.437,0.8312,0.0686
26+2018,ml_theory,13524,true,6.886,2.0,0.3478,1.312,0.7876,0.1378
27+2018,multimodal_ai,470,true,8.711,3.0,0.2723,1.405,0.7541,0.2106
28+2018,nlp,23355,true,7.072,1.0,0.4038,1.289,0.832,0.1545
29+2018,operating_systems,1555,true,3.143,0.0,0.5498,0.667,0.8349,0.11
30+2018,privacy_preserving_computation,3104,true,5.156,1.0,0.3785,1.111,0.7674,0.2146
31+2018,programming_languages,4389,true,2.433,0.0,0.5459,0.631,0.806,0.067
32+2018,quantum_computing,6224,true,6.366,1.0,0.4015,1.237,0.7758,0.1496
33+2018,reinforcement_learning,3296,true,8.653,2.0,0.3547,1.556,0.8069,0.1462
34+2018,robotics,17489,true,5.921,2.0,0.2809,1.07,0.7157,0.233
35+2018,scientific_computing,16874,true,5.02,2.0,0.3062,0.834,0.7039,0.1582
36+2018,software_engineering,29000,true,2.259,0.0,0.6034,0.992,0.8387,0.1127
37+2018,speech_audio,12081,true,4.069,1.0,0.4724,0.9,0.8142,0.1103
38+2018,sustainable_computing,583,true,4.76,2.0,0.3499,1.299,0.7194,0.2401
39+2019,ai_accelerators,992,true,7.205,2.0,0.2843,1.182,0.736,0.1935
40+2019,ai_agents,5505,true,3.783,1.0,0.4534,0.69,0.7839,0.1006
41+2019,ai_safety,1477,true,11.343,2.0,0.3764,1.388,0.8231,0.3236
42+2019,algorithms,29232,true,4.506,1.0,0.3759,0.874,0.7533,0.125
43+2019,applied_cryptography,11344,true,4.055,1.0,0.4324,0.777,0.7792,0.0896
44+2019,blockchain,9040,true,7.389,1.0,0.4887,2.591,0.8618,0.2094
45+2019,cloud_computing,8737,true,3.471,0.0,0.517,1.116,0.8117,0.1444
46+2019,compilers,689,true,3.091,1.0,0.4862,0.689,0.7958,0.09
47+2019,computational_biology,14860,true,10.899,3.0,0.253,1.458,0.7621,0.2201
48+2019,computational_complexity,3947,true,3.656,1.0,0.4697,0.691,0.8121,0.0877
49+2019,computer_architecture,18836,true,3.591,1.0,0.4137,0.862,0.7504,0.1117
50+2019,computer_graphics,9793,true,2.737,0.0,0.5835,0.899,0.8426,0.1103
51+2019,computer_networks,58190,true,6.023,2.0,0.3408,1.086,0.7592,0.1507
52+2019,computer_vision,81509,true,8.555,2.0,0.3508,1.235,0.8068,0.1635
53+2019,cybersecurity,31518,true,4.269,1.0,0.4827,1.051,0.8122,0.1412
54+2019,database_systems,16547,true,3.087,0.0,0.5146,0.838,0.812,0.1169
55+2019,deep_learning_architectures,15335,true,9.466,2.0,0.3042,1.441,0.765,0.2072
56+2019,distributed_computing,11480,true,4.426,1.0,0.4091,1.133,0.7744,0.1264
57+2019,edge_computing,6643,true,8.101,2.0,0.3372,1.932,0.7738,0.2585
58+2019,federated_learning,215,true,6.674,2.0,0.3581,1.096,0.7523,0.1814
59+2019,formal_verification,4799,true,3.43,1.0,0.3755,0.876,0.7127,0.0904
60+2019,generative_ai,1266,true,7.974,2.0,0.3262,1.274,0.7584,0.1564
61+2019,hci,37191,true,3.595,1.0,0.4882,0.928,0.7961,0.1612
62+2019,information_retrieval,12284,true,4.71,1.0,0.4365,1.428,0.7956,0.1785
63+2019,llm,1857,true,4.361,0.0,0.5644,0.68,0.8747,0.0797
64+2019,ml_theory,17048,true,7.794,2.0,0.3277,1.311,0.7821,0.1282
65+2019,multimodal_ai,632,true,9.858,2.0,0.3054,1.449,0.7606,0.2089
66+2019,nlp,25718,true,8.644,1.0,0.3792,1.345,0.8354,0.156
67+2019,operating_systems,1578,true,3.187,0.0,0.5247,0.913,0.8151,0.0963
68+2019,privacy_preserving_computation,3249,true,6.127,2.0,0.3573,1.114,0.7752,0.1961
69+2019,programming_languages,4473,true,2.702,0.0,0.5245,0.609,0.7922,0.0753
70+2019,quantum_computing,7481,true,7.34,1.0,0.3886,1.147,0.7899,0.1269
71+2019,reinforcement_learning,4740,true,7.097,2.0,0.3302,1.171,0.7619,0.127
72+2019,robotics,20051,true,6.552,2.0,0.2618,1.084,0.7064,0.2315
73+2019,scientific_computing,18317,true,5.681,2.0,0.2879,0.88,0.7077,0.1512
74+2019,software_engineering,28927,true,2.55,0.0,0.5871,0.97,0.8365,0.1106
75+2019,speech_audio,12772,true,4.816,1.0,0.441,0.94,0.8033,0.1179
76+2019,sustainable_computing,664,true,4.88,1.0,0.3479,0.801,0.7247,0.1943
77+2020,ai_accelerators,1083,true,9.403,3.0,0.2355,1.252,0.712,0.2068
78+2020,ai_agents,5128,true,4.697,1.0,0.4142,0.755,0.7831,0.1047
79+2020,ai_safety,2153,true,10.482,2.0,0.3525,1.37,0.7983,0.3065
80+2020,algorithms,29331,true,4.937,1.0,0.3607,0.805,0.758,0.1167
81+2020,applied_cryptography,11260,true,4.715,1.0,0.4091,0.8,0.7799,0.0948
82+2020,blockchain,9677,true,7.334,1.0,0.4284,2.168,0.8169,0.2259
83+2020,cloud_computing,8142,true,3.913,1.0,0.479,1.014,0.8001,0.1393
84+2020,compilers,735,true,3.078,1.0,0.4803,0.685,0.7732,0.0912
85+2020,computational_biology,18059,true,12.517,4.0,0.2407,1.556,0.7571,0.203
86+2020,computational_complexity,3870,true,3.544,1.0,0.4488,0.553,0.7827,0.0817
87+2020,computer_architecture,17366,true,4.415,1.0,0.3912,0.896,0.7658,0.1089
88+2020,computer_graphics,9432,true,3.086,0.0,0.5322,0.918,0.8173,0.111
89+2020,computer_networks,54158,true,7.381,2.0,0.3036,1.153,0.7523,0.1577
90+2020,computer_vision,83704,true,9.266,2.0,0.3163,1.223,0.7869,0.1595
91+2020,cybersecurity,32637,true,4.593,1.0,0.4547,1.001,0.7945,0.1378
92+2020,database_systems,15010,true,3.538,1.0,0.4758,0.828,0.796,0.1152
93+2020,deep_learning_architectures,17859,true,9.962,3.0,0.2752,1.378,0.7535,0.1942
94+2020,distributed_computing,10877,true,4.648,1.0,0.3762,0.981,0.7522,0.1199
95+2020,edge_computing,7076,true,9.313,2.0,0.3005,1.928,0.7645,0.2431
96+2020,federated_learning,304,true,13.549,3.0,0.2961,1.957,0.8065,0.1908
97+2020,formal_verification,4535,true,3.228,1.0,0.3693,0.783,0.7083,0.0776
98+2020,generative_ai,1497,true,7.874,2.0,0.3213,1.298,0.7809,0.1403
99+2020,hci,36212,true,4.191,1.0,0.452,0.918,0.7839,0.154
100+2020,information_retrieval,11744,true,5.353,1.0,0.3964,1.379,0.7806,0.1782
101+2020,llm,2460,true,5.87,1.0,0.4996,0.966,0.8538,0.0915
102+2020,ml_theory,19283,true,7.812,2.0,0.3118,1.313,0.7794,0.1217
103+2020,multimodal_ai,755,true,10.449,3.0,0.245,1.474,0.7257,0.2079
104+2020,nlp,28702,true,7.807,2.0,0.3478,1.221,0.7997,0.1513
105+2020,operating_systems,1453,true,3.581,1.0,0.49,0.66,0.7996,0.0957
106+2020,privacy_preserving_computation,4153,true,7.689,2.0,0.3311,1.482,0.7819,0.1813
107+2020,programming_languages,4125,true,2.557,0.0,0.5205,0.551,0.789,0.0562
108+2020,quantum_computing,8095,true,7.636,2.0,0.3548,1.039,0.763,0.1193
109+2020,reinforcement_learning,5671,true,7.154,2.0,0.3229,1.161,0.7587,0.1344
110+2020,robotics,20335,true,7.423,3.0,0.2297,1.074,0.6884,0.222
111+2020,scientific_computing,19721,true,6.64,2.0,0.2715,0.871,0.7289,0.1255
112+2020,software_engineering,27734,true,2.641,0.0,0.5665,0.897,0.8206,0.1061
113+2020,speech_audio,12456,true,5.207,1.0,0.4035,0.952,0.7782,0.113
114+2020,sustainable_computing,632,true,7.161,3.0,0.2547,1.134,0.7132,0.2373
115+2021,ai_accelerators,1248,true,8.878,3.0,0.2508,1.322,0.7274,0.1947
116+2021,ai_agents,4891,true,4.554,1.0,0.4146,0.732,0.7715,0.0945
117+2021,ai_safety,2493,true,10.485,3.0,0.3097,1.199,0.7706,0.3277
118+2021,algorithms,28947,true,5.145,1.0,0.3543,0.847,0.7668,0.1074
119+2021,applied_cryptography,10900,true,4.666,1.0,0.3814,0.767,0.7552,0.0948
120+2021,autonomous_scientific_discovery,221,true,13.457,4.0,0.276,1.775,0.7559,0.2127
121+2021,blockchain,10604,true,7.152,1.0,0.3858,1.939,0.794,0.2252
122+2021,cloud_computing,7665,true,4.034,1.0,0.4515,0.977,0.7918,0.1335
123+2021,compilers,708,true,4.056,1.0,0.4506,0.803,0.7879,0.0946
124+2021,computational_biology,19175,true,10.837,4.0,0.2233,1.331,0.7304,0.198
125+2021,computational_complexity,3762,true,3.8,1.0,0.4285,0.639,0.7844,0.0904
126+2021,computer_architecture,16322,true,3.914,1.0,0.3597,0.894,0.7197,0.1116
127+2021,computer_graphics,8918,true,3.687,0.0,0.5038,0.902,0.8281,0.1092
128+2021,computer_networks,51370,true,6.978,2.0,0.2777,1.082,0.7311,0.1519
129+2021,computer_vision,92549,true,9.505,2.0,0.2779,1.224,0.7729,0.1638
130+2021,cybersecurity,31638,true,4.976,1.0,0.4002,1.046,0.7721,0.1464
131+2021,database_systems,14240,true,3.84,1.0,0.432,0.814,0.7846,0.1094
132+2021,deep_learning_architectures,21101,true,9.087,3.0,0.2758,1.284,0.7493,0.1855
133+2021,distributed_computing,10070,true,4.81,1.0,0.3547,0.979,0.7419,0.1226
134+2021,edge_computing,7817,true,8.715,2.0,0.3019,1.622,0.7592,0.2328
135+2021,federated_learning,651,true,11.476,3.0,0.2273,1.844,0.7565,0.2258
136+2021,formal_verification,4509,true,3.442,1.0,0.3728,0.877,0.7276,0.0854
137+2021,foundation_models,249,true,13.245,4.0,0.2209,1.544,0.7219,0.2892
138+2021,generative_ai,1956,true,7.059,2.0,0.3359,1.135,0.7652,0.1493
139+2021,hci,34354,true,4.851,1.0,0.3901,0.954,0.7632,0.1681
140+2021,information_retrieval,11478,true,5.664,1.0,0.3643,1.414,0.7676,0.1847
141+2021,llm,2844,true,8.016,1.0,0.3966,1.539,0.8451,0.1055
142+2021,ml_theory,19921,true,6.804,2.0,0.3231,1.121,0.7754,0.12
143+2021,multimodal_ai,948,true,9.047,4.0,0.2384,1.22,0.6904,0.2141
144+2021,nlp,29505,true,7.983,2.0,0.323,1.247,0.7918,0.149
145+2021,operating_systems,1320,true,3.407,1.0,0.4348,0.714,0.7505,0.1068
146+2021,privacy_preserving_computation,4235,true,6.791,2.0,0.3159,1.186,0.7568,0.179
147+2021,programming_languages,3975,true,2.758,1.0,0.4989,0.61,0.7904,0.0652
148+2021,quantum_computing,9279,true,7.724,2.0,0.3583,1.003,0.7701,0.108
149+2021,reinforcement_learning,6584,true,6.362,2.0,0.3182,1.048,0.756,0.1303
150+2021,robotics,22041,true,7.414,3.0,0.223,1.066,0.6881,0.2129
151+2021,scientific_computing,20255,true,6.095,2.0,0.2856,0.934,0.7267,0.1155
152+2021,software_engineering,24646,true,3.179,1.0,0.4945,0.967,0.7967,0.1191
153+2021,speech_audio,12716,true,5.412,2.0,0.3528,0.96,0.7551,0.1194
154+2021,sustainable_computing,712,true,7.677,3.0,0.2725,1.115,0.7153,0.2135
155+2022,ai_accelerators,1419,true,11.144,4.0,0.2114,1.538,0.7266,0.203
156+2022,ai_agents,4514,true,5.105,1.0,0.3881,0.798,0.7749,0.1094
157+2022,ai_safety,2954,true,10.3,3.0,0.2576,1.301,0.7371,0.3544
158+2022,algorithms,28070,true,5.326,1.0,0.3383,0.84,0.7591,0.1116
159+2022,applied_cryptography,10252,true,5.086,2.0,0.3447,0.849,0.7438,0.1036
160+2022,autonomous_scientific_discovery,236,true,12.182,4.0,0.2034,1.574,0.7494,0.2415
161+2022,blockchain,11670,true,7.813,1.0,0.3642,1.93,0.8068,0.2324
162+2022,cloud_computing,7024,true,4.064,1.0,0.4311,0.944,0.7795,0.1399
163+2022,compilers,642,true,3.104,1.0,0.4517,0.583,0.7645,0.0981
164+2022,computational_biology,19473,true,10.379,4.0,0.2188,1.26,0.7348,0.1878
165+2022,computational_complexity,3534,true,4.083,1.0,0.416,0.675,0.7721,0.1047
166+2022,computer_architecture,15145,true,4.234,2.0,0.3098,0.916,0.6897,0.1269
167+2022,computer_graphics,8211,true,3.862,1.0,0.4833,0.862,0.8198,0.1125
168+2022,computer_networks,48795,true,7.373,3.0,0.2485,1.111,0.7194,0.1596
169+2022,computer_vision,97846,true,9.471,3.0,0.2604,1.205,0.76,0.1607
170+2022,cybersecurity,31362,true,5.428,1.0,0.3646,1.047,0.7601,0.1551
171+2022,database_systems,13464,true,3.681,1.0,0.4302,0.734,0.7781,0.1114
172+2022,deep_learning_architectures,22083,true,8.706,3.0,0.2548,1.237,0.7271,0.1804
173+2022,distributed_computing,9356,true,4.899,2.0,0.3361,0.912,0.7306,0.129
174+2022,edge_computing,8423,true,8.55,2.0,0.2779,1.549,0.7501,0.2224
175+2022,federated_learning,973,true,11.104,4.0,0.2261,1.674,0.7366,0.2035
176+2022,formal_verification,3988,true,3.15,1.0,0.3691,0.842,0.7028,0.0905
177+2022,foundation_models,350,true,9.843,3.0,0.2086,1.271,0.6979,0.2429
178+2022,generative_ai,2149,true,9.292,2.0,0.2946,1.36,0.7849,0.1615
179+2022,hci,34678,true,5.535,1.0,0.3641,0.996,0.7667,0.1798
180+2022,information_retrieval,11090,true,5.78,2.0,0.3409,1.428,0.7562,0.2022
181+2022,llm,3255,true,11.527,2.0,0.3386,2.447,0.8643,0.122
182+2022,ml_theory,20026,true,6.707,2.0,0.3188,1.07,0.7702,0.1255
183+2022,multimodal_ai,1276,true,9.703,3.0,0.2147,1.421,0.7123,0.2108
184+2022,nlp,29453,true,6.499,2.0,0.3163,1.17,0.7648,0.1432
185+2022,operating_systems,1080,true,3.231,1.0,0.4361,0.631,0.7445,0.1019
186+2022,privacy_preserving_computation,4358,true,6.766,2.0,0.2907,1.039,0.7333,0.1824
187+2022,programming_languages,3658,true,2.709,1.0,0.4956,0.634,0.7815,0.0637
188+2022,quantum_computing,9202,true,7.688,2.0,0.3057,0.936,0.743,0.1047
189+2022,reinforcement_learning,7002,true,6.362,2.0,0.2961,0.937,0.7409,0.1377
190+2022,robotics,23037,true,7.898,3.0,0.2085,1.059,0.6877,0.2125
191+2022,scientific_computing,19884,true,5.65,2.0,0.2547,0.763,0.6808,0.1182
192+2022,software_engineering,23682,true,3.414,1.0,0.4811,0.999,0.7958,0.1308
193+2022,speech_audio,12558,true,5.025,2.0,0.3292,0.93,0.7339,0.1181
194+2022,sustainable_computing,825,true,9.013,3.0,0.2364,1.266,0.7097,0.2473
195+2023,ai_accelerators,1599,true,9.201,3.0,0.2395,1.18,0.7326,0.1801
196+2023,ai_agents,4985,true,4.676,1.0,0.4144,0.725,0.7879,0.0931
197+2023,ai_safety,4007,true,11.656,2.0,0.2962,1.489,0.7986,0.29
198+2023,algorithms,28420,true,5.116,1.0,0.3763,0.797,0.7783,0.1099
199+2023,applied_cryptography,11160,true,4.915,1.0,0.3588,0.819,0.7487,0.0978
200+2023,autonomous_scientific_discovery,399,true,24.263,3.0,0.2331,2.467,0.8619,0.2155
201+2023,blockchain,14220,true,5.544,1.0,0.393,1.382,0.7851,0.1954
202+2023,cloud_computing,7556,true,3.962,1.0,0.4465,0.975,0.7889,0.1341
203+2023,compilers,692,true,2.902,1.0,0.4509,0.496,0.7702,0.0896
204+2023,computational_biology,20497,true,9.826,3.0,0.2323,1.311,0.7307,0.179
205+2023,computational_complexity,3492,true,3.78,1.0,0.4674,0.577,0.7993,0.0971
206+2023,computer_architecture,16349,true,4.405,2.0,0.3215,0.863,0.7029,0.1284
207+2023,computer_graphics,9203,true,3.918,1.0,0.4955,0.882,0.8298,0.1206
208+2023,computer_networks,51542,true,7.21,2.0,0.2687,1.058,0.7273,0.1527
209+2023,computer_vision,113444,true,9.361,2.0,0.2817,1.193,0.7723,0.158
210+2023,cybersecurity,36289,true,5.298,1.0,0.3882,1.054,0.7768,0.1477
211+2023,database_systems,14384,true,3.533,1.0,0.4895,0.688,0.811,0.1045
212+2023,deep_learning_architectures,25474,true,8.276,2.0,0.292,1.125,0.7513,0.1698
213+2023,distributed_computing,9777,true,6.507,1.0,0.3543,1.143,0.8034,0.1283
214+2023,edge_computing,8975,true,8.556,2.0,0.2944,1.521,0.7623,0.2139
215+2023,federated_learning,1494,true,8.68,3.0,0.2443,1.195,0.719,0.1867
216+2023,formal_verification,4382,true,19.317,1.0,0.3914,2.284,0.9568,0.0851
217+2023,foundation_models,570,true,11.47,3.0,0.2789,1.863,0.7925,0.1982
218+2023,generative_ai,3631,true,9.174,2.0,0.3087,1.585,0.7998,0.171
219+2023,hci,40447,true,5.164,1.0,0.395,0.956,0.7774,0.1743
220+2023,information_retrieval,12417,true,5.336,1.0,0.3728,1.331,0.7707,0.1915
221+2023,llm,10361,true,19.464,2.0,0.3057,3.841,0.8747,0.1896
222+2023,ml_theory,22810,true,6.154,1.0,0.3615,0.927,0.7928,0.1203
223+2023,multimodal_ai,1925,true,10.985,3.0,0.2374,1.647,0.7561,0.2073
224+2023,nlp,33775,true,5.714,1.0,0.3498,1.008,0.7827,0.1294
225+2023,operating_systems,1169,true,3.694,1.0,0.4636,0.676,0.7716,0.1172
226+2023,privacy_preserving_computation,4989,true,7.043,2.0,0.2963,1.01,0.7478,0.1758
227+2023,programming_languages,3871,true,2.552,0.0,0.536,0.511,0.7961,0.0568
228+2023,quantum_computing,10530,true,7.585,2.0,0.3119,0.866,0.7444,0.1088
229+2023,reinforcement_learning,8407,true,6.385,2.0,0.3304,0.909,0.7619,0.1392
230+2023,robotics,25682,true,7.932,3.0,0.24,1.052,0.7122,0.2085
231+2023,scientific_computing,20381,true,5.443,2.0,0.2841,0.72,0.696,0.1185
232+2023,software_engineering,26205,true,3.346,0.0,0.5188,1.038,0.8232,0.1305
233+2023,speech_audio,14551,true,4.366,1.0,0.3728,0.808,0.7477,0.1114
234+2023,sustainable_computing,1072,true,8.33,2.0,0.2696,1.236,0.7239,0.2397
235+2024,ai_accelerators,1718,false,7.692,2.0,0.2753,1.159,0.7337,0.181
236+2024,ai_agents,5891,false,4.363,1.0,0.4655,0.878,0.8092,0.101
237+2024,ai_safety,6208,false,9.756,2.0,0.3381,1.634,0.7989,0.2932
238+2024,algorithms,29917,false,4.217,1.0,0.4298,0.768,0.7946,0.1082
239+2024,applied_cryptography,12399,false,4.057,1.0,0.4129,0.824,0.7627,0.1062
240+2024,autonomous_scientific_discovery,454,false,11.467,2.0,0.2885,1.603,0.8027,0.1982
241+2024,blockchain,15343,false,4.241,1.0,0.4438,1.279,0.7981,0.1766
242+2024,cloud_computing,8589,false,3.594,1.0,0.4932,0.983,0.8085,0.1373
243+2024,compilers,768,false,2.896,0.0,0.5234,0.659,0.8197,0.0859
244+2024,computational_biology,23470,false,7.75,3.0,0.2687,1.319,0.7273,0.1755
245+2024,computational_complexity,3819,false,3.479,1.0,0.4836,0.631,0.7984,0.1194
246+2024,computer_architecture,17140,false,3.781,1.0,0.3568,0.958,0.7173,0.1307
247+2024,computer_graphics,10590,false,3.007,0.0,0.5437,0.792,0.8325,0.121
248+2024,computer_networks,54093,false,6.265,2.0,0.308,1.065,0.7392,0.1546
249+2024,computer_vision,132887,false,7.414,2.0,0.3355,1.161,0.7865,0.1498
250+2024,cybersecurity,45440,false,4.314,1.0,0.4422,1.036,0.795,0.1347
251+2024,database_systems,14983,false,2.834,0.0,0.5443,0.633,0.8331,0.0925
252+2024,deep_learning_architectures,28918,false,6.753,2.0,0.3405,1.129,0.7601,0.1638
253+2024,distributed_computing,10586,false,4.18,1.0,0.4036,0.912,0.7644,0.1205
254+2024,edge_computing,10172,false,8.301,2.0,0.3364,1.533,0.8043,0.1996
255+2024,federated_learning,2068,false,5.487,2.0,0.3235,0.992,0.7262,0.1426
256+2024,formal_verification,4233,false,2.313,1.0,0.4399,0.761,0.7352,0.0761
257+2024,foundation_models,1041,false,9.522,1.0,0.3497,1.854,0.8165,0.1844
258+2024,generative_ai,7132,false,7.88,1.0,0.3731,1.837,0.8175,0.1977
259+2024,hci,48038,false,4.404,1.0,0.4477,1.04,0.7956,0.1733
260+2024,information_retrieval,14101,false,4.729,1.0,0.4208,1.437,0.7953,0.1857
261+2024,llm,24859,false,7.931,1.0,0.3812,2.23,0.8322,0.182
262+2024,ml_theory,26665,false,5.826,1.0,0.4254,1.115,0.8278,0.1206
263+2024,multimodal_ai,3407,false,8.14,2.0,0.3038,1.414,0.7807,0.187
264+2024,nlp,38743,false,4.42,1.0,0.4261,0.979,0.8093,0.1229
265+2024,operating_systems,1259,false,3.152,1.0,0.471,0.68,0.779,0.116
266+2024,privacy_preserving_computation,6674,false,5.44,1.0,0.356,0.983,0.7656,0.1552
267+2024,programming_languages,3845,false,2.181,0.0,0.5873,0.696,0.8262,0.0572
268+2024,quantum_computing,12711,false,5.883,2.0,0.356,0.929,0.7512,0.1063
269+2024,reinforcement_learning,9881,false,5.27,1.0,0.3867,0.826,0.771,0.1245
270+2024,robotics,29944,false,6.481,2.0,0.2777,0.996,0.7176,0.1962
271+2024,scientific_computing,21877,false,4.549,2.0,0.3194,0.733,0.702,0.1145
272+2024,software_engineering,30519,false,3.007,0.0,0.5638,1.107,0.8428,0.1268
273+2024,speech_audio,16395,false,3.627,1.0,0.4342,0.801,0.7826,0.1045
274+2024,sustainable_computing,1269,false,6.861,2.0,0.2971,1.266,0.7547,0.2285
275+2025,ai_accelerators,1906,false,3.322,1.0,0.4717,1.273,0.7923,0.1663
276+2025,ai_agents,8818,false,1.874,0.0,0.6702,0.983,0.8825,0.1034
277+2025,ai_safety,10883,false,3.068,0.0,0.5784,1.417,0.859,0.2193
278+2025,algorithms,33790,false,1.623,0.0,0.626,0.79,0.844,0.1037
279+2025,applied_cryptography,14611,false,1.508,0.0,0.6241,0.824,0.8327,0.0986
280+2025,autonomous_scientific_discovery,764,false,4.135,1.0,0.4948,1.81,0.8148,0.1767
281+2025,blockchain,18004,false,1.394,0.0,0.6738,1.194,0.8613,0.1515
282+2025,cloud_computing,10971,false,1.238,0.0,0.6969,0.957,0.8711,0.1149
283+2025,compilers,767,false,1.14,0.0,0.6923,0.674,0.8664,0.0847
284+2025,computational_biology,28138,false,3.216,1.0,0.4504,1.386,0.7771,0.1611
285+2025,computational_complexity,4208,false,1.278,0.0,0.6761,0.602,0.8533,0.1008
286+2025,computer_architecture,20207,false,1.366,0.0,0.5846,0.964,0.7929,0.116
287+2025,computer_graphics,12438,false,0.925,0.0,0.7087,0.82,0.8516,0.0996
288+2025,computer_networks,59543,false,2.371,0.0,0.5238,1.02,0.8013,0.1346
289+2025,computer_vision,150699,false,2.364,0.0,0.5485,1.104,0.8144,0.1317
290+2025,cybersecurity,59022,false,1.496,0.0,0.6617,1.058,0.8595,0.1175
291+2025,database_systems,16680,false,1.138,0.0,0.7238,0.67,0.8852,0.0932
292+2025,deep_learning_architectures,34123,false,2.476,0.0,0.5597,1.075,0.8225,0.1404
293+2025,distributed_computing,12549,false,1.456,0.0,0.6247,0.953,0.8302,0.1094
294+2025,edge_computing,11545,false,2.578,0.0,0.5509,1.424,0.8198,0.1778
295+2025,federated_learning,2543,false,2.114,0.0,0.5395,1.015,0.7978,0.1298
296+2025,formal_verification,4623,false,0.81,0.0,0.6652,0.668,0.8125,0.0774
297+2025,foundation_models,1854,false,2.303,0.0,0.6149,1.205,0.8601,0.1176
298+2025,generative_ai,12739,false,2.584,0.0,0.6143,1.68,0.8668,0.1675
299+2025,hci,58127,false,1.578,0.0,0.6506,1.047,0.8498,0.1546
300+2025,information_retrieval,17369,false,1.556,0.0,0.6342,1.246,0.8396,0.1692
301+2025,llm,45150,false,2.147,0.0,0.6603,1.309,0.8883,0.1345
302+2025,ml_theory,32789,false,1.865,0.0,0.6384,0.888,0.8578,0.109
303+2025,multimodal_ai,6746,false,2.042,0.0,0.5913,1.162,0.8314,0.1453
304+2025,nlp,42900,false,1.981,0.0,0.6565,1.531,0.8975,0.1131
305+2025,operating_systems,1363,false,1.116,0.0,0.6698,0.646,0.8337,0.1042
306+2025,privacy_preserving_computation,8811,false,2.118,0.0,0.5932,1.144,0.8456,0.1338
307+2025,programming_languages,4193,false,0.717,0.0,0.7412,0.465,0.8646,0.0501
308+2025,quantum_computing,17471,false,2.102,0.0,0.5806,0.888,0.82,0.1051
309+2025,reinforcement_learning,13207,false,2.103,0.0,0.6116,0.916,0.8431,0.1145
310+2025,robotics,33462,false,2.341,0.0,0.513,0.986,0.785,0.1632
311+2025,scientific_computing,23186,false,1.869,0.0,0.5261,0.756,0.774,0.1158
312+2025,software_engineering,36381,false,1.597,0.0,0.7281,1.47,0.9227,0.1175
313+2025,speech_audio,18367,false,1.166,0.0,0.6498,0.726,0.8281,0.0985
314+2025,sustainable_computing,1639,false,2.614,0.0,0.5101,1.437,0.7938,0.1989
315+2026,ai_accelerators,1649,false,0.415,0.0,0.8441,1.358,0.9238,0.1116
316+2026,ai_agents,12209,false,0.197,0.0,0.9199,1.145,0.9597,0.0799
317+2026,ai_safety,12544,false,0.308,0.0,0.8858,1.364,0.944,0.1193
318+2026,algorithms,28772,false,0.194,0.0,0.9081,0.778,0.9482,0.0722
319+2026,applied_cryptography,12178,false,0.197,0.0,0.9079,0.888,0.95,0.0852
320+2026,autonomous_scientific_discovery,910,false,0.556,0.0,0.8418,1.937,0.9308,0.1044
321+2026,blockchain,13274,false,0.15,0.0,0.9251,1.193,0.9552,0.0971
322+2026,cloud_computing,8774,false,0.128,0.0,0.9389,0.965,0.9647,0.0782
323+2026,compilers,621,false,0.116,0.0,0.9147,0.793,0.9339,0.0886
324+2026,computational_biology,23779,false,0.411,0.0,0.8212,1.314,0.9012,0.1294
325+2026,computational_complexity,4067,false,0.16,0.0,0.9169,0.628,0.9471,0.0725
326+2026,computer_architecture,15793,false,0.138,0.0,0.9175,0.9,0.9454,0.0795
327+2026,computer_graphics,10260,false,0.128,0.0,0.9347,0.971,0.9642,0.0748
328+2026,computer_networks,42415,false,0.288,0.0,0.8715,1.008,0.9287,0.1073
329+2026,computer_vision,117907,false,0.25,0.0,0.8876,0.978,0.938,0.0913
330+2026,cybersecurity,51618,false,0.169,0.0,0.9232,1.145,0.9586,0.088
331+2026,database_systems,14095,false,0.104,0.0,0.9445,0.642,0.9664,0.0817
332+2026,deep_learning_architectures,26151,false,2.034,0.0,0.8777,2.99,0.9909,0.0972
333+2026,distributed_computing,10002,false,0.182,0.0,0.9151,1.008,0.9537,0.0797
334+2026,edge_computing,8717,false,0.24,0.0,0.8878,1.02,0.9347,0.1185
335+2026,federated_learning,2841,false,0.214,0.0,0.8976,0.748,0.9396,0.0982
336+2026,formal_verification,4866,false,0.229,0.0,0.9231,2.011,0.9712,0.0629
337+2026,foundation_models,2688,false,0.298,0.0,0.9159,1.282,0.9694,0.0632
338+2026,generative_ai,14510,false,0.272,0.0,0.897,1.563,0.9481,0.1239
339+2026,hci,47218,false,0.155,0.0,0.9151,0.98,0.9462,0.0971
340+2026,information_retrieval,16901,false,0.172,0.0,0.9172,1.169,0.9533,0.1021
341+2026,llm,51454,false,0.335,0.0,0.925,1.135,0.9788,0.0757
342+2026,ml_theory,29457,false,0.241,0.0,0.9048,1.011,0.953,0.0836
343+2026,multimodal_ai,8806,false,0.224,0.0,0.9068,1.185,0.9498,0.098
344+2026,nlp,34783,false,0.136,0.0,0.9295,0.669,0.9577,0.0687
345+2026,operating_systems,1247,false,0.142,0.0,0.9326,1.56,0.9633,0.0706
346+2026,privacy_preserving_computation,7623,false,0.213,0.0,0.8974,0.958,0.9423,0.0938
347+2026,programming_languages,3691,false,0.15,0.0,0.9176,1.386,0.9483,0.0718
348+2026,quantum_computing,19271,false,0.203,0.0,0.8961,0.658,0.936,0.0706
349+2026,reinforcement_learning,12554,false,0.204,0.0,0.9043,0.706,0.9453,0.0819
350+2026,robotics,27671,false,0.233,0.0,0.8899,0.789,0.9364,0.1103
351+2026,scientific_computing,18737,false,0.256,0.0,0.8673,0.8,0.918,0.0886
352+2026,software_engineering,34261,false,0.139,0.0,0.9333,1.311,0.9628,0.0893
353+2026,speech_audio,15624,false,0.148,0.0,0.9406,0.703,0.9739,0.0586
354+2026,sustainable_computing,1328,false,0.349,0.0,0.8494,2.162,0.9188,0.1401
modified results/bibliometrics/category_share_ci_yearly.csv +320 −320
@@ -1,321 +1,321 @@
1 year,category,share,ci_low,ci_high 1 year,category,share,ci_low,ci_high
2−2018,ai_accelerators,0.001637437670274501,0.0015132286147237274,0.0017456326654802915 2+2018,ai_accelerators,0.002409752131363896,0.0021650945341636866,0.0026720851720704424
3−2018,ai_agents,0.010881554004415877,0.010604260540754784,0.011113757879418129 3+2018,ai_agents,0.016013951816741003,0.015468811679933069,0.016632592339462786
4−2018,ai_safety,0.0021058489353652505,0.0019826908025246167,0.0022380650210885155 4+2018,ai_safety,0.0030990944281110255,0.0028395599992929823,0.003373211097821796
5−2018,algorithms,0.058347228866945176,0.05766232752356569,0.05901791773176849 5+2018,algorithms,0.05724486970252229,0.0561930335185327,0.05852059483759213
6−2018,applied_cryptography,0.021582949829950696,0.021211573761762827,0.02194957172397365 6+2018,applied_cryptography,0.0211751813205913,0.020432518102600057,0.021796915045926693
7−2018,autonomous_scientific_discovery,0.00017215114870856613,0.00013907210530846083,0.00021123546473803995 7+2018,autonomous_scientific_discovery,0.0002533480235908253,0.00017675443506336646,0.0003358334266203963
8−2018,blockchain,0.012893320335254354,0.01257243860109176,0.013200740249939447 8+2018,blockchain,0.012649725736034926,0.012084111543832154,0.0132153399282377
9−2018,cloud_computing,0.01880851387518241,0.01841101487105679,0.019190949653796033 9+2018,cloud_computing,0.01845316302061546,0.01784041431239579,0.019110247632963522
10−2018,compilers,0.0014512742187640748,0.001354138933984038,0.0015513620959202179 10+2018,compilers,0.0014238551713437854,0.0012431728599456775,0.0016025735445745226
11−2018,computational_biology,0.025958791819217272,0.02551665362188001,0.026400329489291597 11+2018,computational_biology,0.025468350154463738,0.02472794546536497,0.02618042503549818
12−2018,computational_complexity,0.007680743692962421,0.0074505415755032915,0.007914949325507795 12+2018,computational_complexity,0.007535630748201524,0.007137785973946396,0.007889139618328257
13−2018,computer_architecture,0.03914236699822444,0.038611801161419726,0.03963489944170982 13+2018,computer_architecture,0.03840284692476742,0.037570922717069175,0.03931607817259481
14−2018,computer_graphics,0.020413923424766944,0.02001547358580834,0.020802464563887093 14+2018,computer_graphics,0.02002824143084679,0.019395853340953412,0.020706929363035934
15−2018,computer_networks,0.1175912451132094,0.11667899415686973,0.11851590696631642 15+2018,computer_networks,0.11536958370402665,0.113782427074066,0.11679712125943426
16−2018,computer_vision,0.13959856754230215,0.13864142717305794,0.14057162188401418 16+2018,computer_vision,0.13696111991610055,0.135278859580385,0.1385552976250096
17−2018,cybersecurity,0.05893174206953705,0.05832616036880381,0.05954638172315293 17+2018,cybersecurity,0.057818339647394544,0.05672211846082238,0.05889516694456392
18−2018,database_systems,0.032558586438893346,0.03206815584082825,0.03305016804754574 18+2018,database_systems,0.03194345429228506,0.031152821884555787,0.0328440672373871
19−2018,deep_learning_architectures,0.02400507645712936,0.023564589709765174,0.024445563204493544 19+2018,deep_learning_architectures,0.023551546503109897,0.02283947162207545,0.02420666717729059
20−2018,distributed_computing,0.023058245139232245,0.022629368585618172,0.02350398650014713 20+2018,distributed_computing,0.022622603749943536,0.021893982689848993,0.023379014535106376
21−2018,edge_computing,0.010379112861092039,0.010085805337085961,0.010655355402042994 21+2018,edge_computing,0.010183019397817279,0.009712508911369435,0.010628833361810437
22−2018,federated_learning,0.00028625132866656923,0.0002442144202609891,0.000341349705041026 22+2018,federated_learning,0.0004212647369010234,0.00032404979761617187,0.000518479676185875
23−2018,formal_verification,0.010623327281353028,0.010356042605407548,0.010894615472384754 23+2018,formal_verification,0.010422619854236509,0.00994228967695181,0.010988823227889492
24−2018,foundation_models,9.208084698365164e-05,6.605799892305444e-05,0.00011810369504424884 24+2018,foundation_models,0.00013551173354858095,8.837721753168323e-05,0.0001885380640675909
25−2018,generative_ai,0.0015333462780321122,0.0014250511949491652,0.0016344850778983947 25+2018,generative_ai,0.0015043766362059857,0.0013225650603616397,0.0016763685212134779
26−2018,hci,0.06851015191337995,0.0678100872566113,0.06915541845740561 26+2018,hci,0.06721578377826352,0.06605593099506855,0.0683425442033383
27−2018,information_retrieval,0.024211257484071016,0.02380680237248304,0.02469873548976001 27+2018,information_retrieval,0.023753832134349083,0.023113293701061115,0.024454123386379305
28−2018,llm,0.003272873583005879,0.003112632891678894,0.0034310124289125853 28+2018,llm,0.0032110389036511576,0.00295179906555822,0.0034584951127398706
29−2018,ml_theory,0.02709178658862481,0.026672268251524838,0.027592626325914152 29+2018,ml_theory,0.026579939157195573,0.025785084282406452,0.027376757970152068
30−2018,multimodal_ai,0.000940826045267745,0.0008627575010859535,0.0010269016196220282 30+2018,multimodal_ai,0.000923050938664247,0.0007836113287809246,0.0010812952565001442
31−2018,nlp,0.04678507729786753,0.04615212156273208,0.04735272569315859 31+2018,nlp,0.045901162847788904,0.04488826173796744,0.046904735251315346
32−2018,operating_systems,0.003114734737099173,0.0029674554258639087,0.003269920990629773 32+2018,operating_systems,0.004583831682643304,0.0042625804969156356,0.004941022936833856
33−2018,privacy_preserving_computation,0.006215457171396486,0.005999267356739217,0.00644165577376937 33+2018,privacy_preserving_computation,0.006098028009686143,0.005709168252546737,0.006472305525932821
34−2018,programming_languages,0.008787715614309364,0.00853634491083171,0.009009910701596001 34+2018,programming_languages,0.008621688554757541,0.008177838528931756,0.009052920277857972
35−2018,quantum_computing,0.012458938948396692,0.012182596319568582,0.012739285092311049 35+2018,quantum_computing,0.018335326730573216,0.017710499802624215,0.018983426325805557
36−2018,reinforcement_learning,0.006597792862132953,0.006364538064420561,0.006824091552382992 36+2018,reinforcement_learning,0.009709710299480931,0.009264731009208906,0.010163380016143572
37−2018,robotics,0.035022749974477595,0.03457415610906376,0.03550672490446612 37+2018,robotics,0.034361062176318444,0.03349776405639645,0.035212576667236206
38−2018,scientific_computing,0.033793670843000154,0.033273113793911055,0.03426913830343041 38+2018,scientific_computing,0.033155204141552805,0.032280957066347725,0.03399105622558579
39−2018,software_engineering,0.05808299687125296,0.05749533090053067,0.05869928797484191 39+2018,software_engineering,0.05698562986442935,0.055954267735834604,0.05796985747700719
40−2018,speech_audio,0.024213259241614136,0.023774674163915918,0.024626772306084743 40+2018,speech_audio,0.023755796072516454,0.023063360573155715,0.02449286206673069
41−2018,sustainable_computing,0.0011670246476406284,0.0010709402855707312,0.0012691142823398945 41+2018,sustainable_computing,0.0017174639273657107,0.0015551444378325194,0.001938406971194919
42−2019,ai_accelerators,0.0018379185572339326,0.0017257814396164086,0.0019417646611986118 42+2019,ai_accelerators,0.0018345044017934501,0.0016421773274118785,0.002011251133851322
43−2019,ai_agents,0.010201189088840759,0.009925038120135399,0.010464370874178541 43+2019,ai_agents,0.0101822391494705,0.009719637056619056,0.01066069897946446
44−2019,ai_safety,0.0027402031715211554,0.00261041870082132,0.002881104085107487 44+2019,ai_safety,0.0027351129135610003,0.0025187449548817326,0.0029736724577458344
45−2019,algorithms,0.05419636455262802,0.05364415525224135,0.05481429982158109 45+2019,algorithms,0.05409568826740121,0.05305952615417049,0.055143454730552544
46−2019,applied_cryptography,0.021037868162692847,0.020641289062717116,0.02140109793400909 46+2019,applied_cryptography,0.02099878778464176,0.020337755931394714,0.02167355068142037
47−2019,autonomous_scientific_discovery,0.00021121241484341565,0.00017601034570284636,0.0002464144839839849 47+2019,autonomous_scientific_discovery,0.0006324601869086289,0.0005270501557571907,0.0007434181144364586
48−2019,blockchain,0.01676174313235422,0.01644112639210288,0.017132383865594795 48+2019,blockchain,0.01673060617240458,0.016130092622130006,0.01733773096586102
49−2019,cloud_computing,0.016222595652359187,0.01587770801180566,0.016573921566084473 49+2019,cloud_computing,0.016192460223894603,0.015625511446697198,0.01676310759867627
50−2019,compilers,0.0012765381914659067,0.0011829284786592088,0.0013776515032209893 50+2019,compilers,0.0019112503016668655,0.0017198478766813593,0.0021111132686263685
51−2019,computational_biology,0.027561367396584657,0.02711017877092902,0.02794400462444024 51+2019,computational_biology,0.02751016883173323,0.026732122597414126,0.028275269974507414
52−2019,computational_complexity,0.007322030381238409,0.007091317872831599,0.007544451875992374 52+2019,computational_complexity,0.007308428826499711,0.0069264099285153554,0.007689661772497413
53−2019,computer_architecture,0.03491674710648256,0.03438111983340157,0.035438525144467437 53+2019,computer_architecture,0.03485188503649129,0.03403634426916174,0.03573136530955875
54−2019,computer_graphics,0.01815315123364725,0.01780279800867453,0.01851276816102538 54+2019,computer_graphics,0.018119429565294578,0.017497787776319915,0.018713331872387287
55−2019,computer_networks,0.10801291730663411,0.10721031013022912,0.10884544624180857 55+2019,computer_networks,0.10781227028241566,0.10634998349500828,0.10915527729773063
56−2019,computer_vision,0.1513133150900154,0.15043252226530873,0.1521450566104854 56+2019,computer_vision,0.15103223235329744,0.1493210761809391,0.15265619408760683
57−2019,cybersecurity,0.058474342323447726,0.05786344746832277,0.05903998399232224 57+2019,cybersecurity,0.058365719178430524,0.05710713265297631,0.05944099396111479
58−2019,database_systems,0.030716584435868315,0.030259744951745375,0.031185420414606264 58+2019,database_systems,0.03065952467473146,0.029935894057370797,0.03136517086133865
59−2019,deep_learning_architectures,0.028463652010871882,0.0280585966231952,0.028890755010273446 59+2019,deep_learning_architectures,0.028410777343500778,0.02762824655961076,0.029118272828900102
60−2019,distributed_computing,0.02129725183004441,0.020941433020652496,0.021700361840215954 60+2019,distributed_computing,0.02125768961554003,0.0205742350144384,0.021947478065004702
61−2019,edge_computing,0.012320724199199246,0.012038042320298068,0.012611789728777321 61+2019,edge_computing,0.012297836967667784,0.011822567178090248,0.012776805354829583
62−2019,federated_learning,0.00040019194391384015,0.00035016795092461013,0.0004623050685421341 62+2019,federated_learning,0.00119834561730056,0.0010374566223852071,0.0013592346122159131
63−2019,formal_verification,0.008896859790158614,0.008633770641844885,0.009142394222414084 63+2019,formal_verification,0.008880332799810632,0.008463177224775796,0.009334474350688078
64−2019,foundation_models,0.00015377745992985526,0.00011945544251780021,0.00018897952907042452 64+2019,foundation_models,0.000153491799746831,9.986213477504667e-05,0.0002108200623028763
65−2019,generative_ai,0.0023455694490505633,0.0022156923413266735,0.002481792192922161 65+2019,generative_ai,0.002341212270837205,0.002130392208534329,0.0025607702699328976
66−2019,hci,0.06894417878204546,0.06825028115336801,0.0695973161201391 66+2019,hci,0.06881610665275995,0.06770916262826043,0.06997852964102336
67−2019,information_retrieval,0.02279611887924023,0.02239310150609274,0.023232670855095314 67+2019,information_retrieval,0.022753772338373595,0.02208035018321928,0.023462331170578395
68−2019,llm,0.0034405390733703756,0.003288521716897549,0.0035888509488810373 68+2019,llm,0.003434147856986327,0.003170484081698322,0.003681167943145158
69−2019,ml_theory,0.03160034164534471,0.031083936554754964,0.03207376315677334 69+2019,ml_theory,0.031541640198577706,0.030632571144836158,0.03241108802569785
70−2019,multimodal_ai,0.001170931984044199,0.0010801477004711518,0.0012654217485794112 70+2019,multimodal_ai,0.001168756836626472,0.001017900287658427,0.0013314951303339557
71−2019,nlp,0.047685834502103785,0.0471130597823771,0.04822322187864179 71+2019,nlp,0.04759725231185465,0.04664934798347837,0.04857211017012624
72−2019,operating_systems,0.002925477219629415,0.0027892544757578174,0.003072816406387508 72+2019,operating_systems,0.0029200427927740503,0.0026851818461734772,0.003167848830919536
73−2019,privacy_preserving_computation,0.006032523006404924,0.005860032867616135,0.006264393477612411 73+2019,privacy_preserving_computation,0.006021316867176888,0.005650393761945314,0.006388541374824201
74−2019,programming_languages,0.008289160912363522,0.008050157390303868,0.008565311881068882 74+2019,programming_languages,0.00827376279599183,0.007886196001631082,0.008679822578273884
75−2019,quantum_computing,0.013862204279459963,0.013575909556620675,0.014174622643082516 75+2019,quantum_computing,0.013836453562720355,0.013309403406963165,0.01446419803770905
76−2019,reinforcement_learning,0.008783842620812575,0.008566099295773342,0.009014555129219386 76+2019,reinforcement_learning,0.008767525573490672,0.008332662962521185,0.00917067271017512
77−2019,robotics,0.03715671034811141,0.0366456782049168,0.03765695027800371 77+2019,robotics,0.037087687276177055,0.03629142544875547,0.03792569702383099
78−2019,scientific_computing,0.03394961657535744,0.03344090035776419,0.034429105811861614 78+2019,scientific_computing,0.03388655106699917,0.033137585056186324,0.0346579860581364
79−2019,software_engineering,0.05360904582012484,0.053033723582236666,0.054117067260037684 79+2019,software_engineering,0.05350946055029584,0.05239849430092345,0.05463013561832692
80−2019,speech_audio,0.02368358156967879,0.02329441343162739,0.024124719078224555 80+2019,speech_audio,0.023639586459804105,0.022995753085323874,0.024347359340022245
81−2019,sustainable_computing,0.0012357779008820898,0.001149579149999722,0.0013312403541698703 81+2019,sustainable_computing,0.0012334822943510396,0.0010818397933963389,0.0014065304288246503
82−2020,ai_accelerators,0.001975957064365208,0.0018529833309354747,0.0020898250048715992 82+2020,ai_accelerators,0.002864752020728307,0.0026244824964091586,0.003081390641282
83−2020,ai_agents,0.009340722380764194,0.009097552180746347,0.00961685555116454 83+2020,ai_agents,0.013542224068493536,0.012992904568113329,0.014120719153969638
84−2020,ai_safety,0.003922775591375722,0.0037433459418212676,0.0040912782894221645 84+2020,ai_safety,0.005687258850367532,0.005354577970541019,0.006006870124200861
85−2020,algorithms,0.05349835458321875,0.05282388850138681,0.05411090127317197 85+2020,algorithms,0.051708113893035174,0.050675350987129296,0.05261163291745289
86−2020,applied_cryptography,0.020540847676474818,0.020194645429529357,0.02089328739157277 86+2020,applied_cryptography,0.019853479595572703,0.019189614019770748,0.02050753196552792
87−2020,autonomous_scientific_discovery,0.0003514836068410001,0.00030595464222429023,0.00040247604721171514 87+2020,autonomous_scientific_discovery,0.0010191652350240797,0.0008765877150984312,0.0011567261385078994
88−2020,blockchain,0.01764884784402141,0.017297273179251175,0.017973105130021618 88+2020,blockchain,0.02558738417776776,0.024866575604810315,0.02635400237981243
89−2020,cloud_computing,0.014904361856926139,0.01457181829936569,0.015266362654593598 89+2020,cloud_computing,0.014405610161376628,0.013856158644774269,0.014960342326865123
90−2020,compilers,0.0013421938769006066,0.001241074046486894,0.0014451348658989875 90+2020,compilers,0.0019459191145407945,0.0017531754301968623,0.0021597853944292675
91−2020,computational_biology,0.03292290489363523,0.03243474333501487,0.033376646555005364 91+2020,computational_biology,0.03182119018785028,0.030902357281662768,0.03264523544653167
92−2020,computational_complexity,0.007051526039836023,0.006846509112166979,0.007273707387165567 92+2020,computational_complexity,0.006815557495705072,0.006442391641084856,0.007158051581378319
93−2020,computer_architecture,0.031853884804434886,0.031391128408070644,0.03230102476593559 93+2020,computer_architecture,0.03078794322246317,0.030081084363646603,0.031549632818880784
94−2020,computer_graphics,0.017200842832192986,0.016871031012509537,0.017515994325269853 94+2020,computer_graphics,0.01662524290984876,0.01601343573098262,0.01723656602923367
95−2020,computer_networks,0.0990509942615293,0.09815539399855401,0.09981319465817765 95+2020,computer_networks,0.0957364040893345,0.09446024727518516,0.09708274952826203
96−2020,computer_vision,0.15306109440703988,0.15208864125179156,0.1539799144419697 96+2020,computer_vision,0.14793913876137096,0.14634024628946404,0.1494509445320641
97−2020,cybersecurity,0.059560991511579835,0.0589581880200546,0.06014230533180599 97+2020,cybersecurity,0.05756787394034978,0.05661330864336609,0.058604905370771954
98−2020,database_systems,0.02742846944369069,0.02702503728822202,0.027842828550667366 98+2020,database_systems,0.026510617624693717,0.025774319148336944,0.027203394753147262
99−2020,deep_learning_architectures,0.032615129092826274,0.03216156954731461,0.033086991282113856 99+2020,deep_learning_architectures,0.03152371363393133,0.0307570074210719,0.03233869377869152
100−2020,distributed_computing,0.019852449731470167,0.019483437473251732,0.02025674693726655 100+2020,distributed_computing,0.019188117835919675,0.018590392387416565,0.019765468780803784
101−2020,edge_computing,0.012957543329915626,0.012675946683761275,0.01326896144789392 101+2020,edge_computing,0.012523938941617144,0.012010703875292194,0.01300623820655082
102−2020,federated_learning,0.0005536322097391919,0.0004971762936144716,0.0006183288684595365 102+2020,federated_learning,0.001605317261385079,0.0014441254541358041,0.0017929123130650292
103−2020,formal_verification,0.008258954181471169,0.008024434484730496,0.008492153538237956 103+2020,formal_verification,0.007982580899540934,0.0076094150449207195,0.008356758878530993
104−2020,foundation_models,0.00024949872609957,0.00021029828756458283,0.00028774305637760633 104+2020,foundation_models,0.0002411496324668375,0.00017148907257723828,0.00030627763539583743
105−2020,generative_ai,0.0027280955598332547,0.0025950599252232286,0.002869371937038905 105+2020,generative_ai,0.002636804010476807,0.002397414594305348,0.0028673923451713744
106−2020,hci,0.06601881985281396,0.06535860433690706,0.06667471011708229 106+2020,hci,0.0638096009237615,0.06280803785169121,0.06500254351254682
107−2020,information_retrieval,0.02144778465163968,0.02108259682644905,0.021837876820475652 107+2020,information_retrieval,0.020730067310671133,0.020134586137944627,0.021334349564874528
108−2020,llm,0.0044818712768689185,0.004304262785899133,0.004667811568363561 108+2020,llm,0.004331892302926184,0.0040396963978933725,0.004641954403357083
109−2020,ml_theory,0.035164751111362025,0.03466211134199355,0.035677999129486195 109+2020,ml_theory,0.03398801644746106,0.03321501746134565,0.034762775649871853
110−2020,multimodal_ai,0.0013749747314246378,0.0012756760595955935,0.0014779157204230188 110+2020,multimodal_ai,0.0019934449545160105,0.0017979289295068576,0.002233714478835159
111−2020,nlp,0.05235284583346233,0.051698093793309426,0.052982192711359115 111+2020,nlp,0.050600937843242176,0.04966423874165657,0.05163246859774128
112−2020,operating_systems,0.0026479645821078455,0.002510421580000765,0.002780044108460921 112+2020,operating_systems,0.0038390317402202377,0.0035775076041343955,0.004148081716281296
113−2020,privacy_preserving_computation,0.007570556236466516,0.007346553730552303,0.007836536447757335 113+2020,privacy_preserving_computation,0.007317219139887909,0.006954614583039963,0.0076702745233334264
114−2020,programming_languages,0.007514100320341795,0.007289050648241398,0.007739149992442192 114+2020,programming_languages,0.007262652434731179,0.0068884744557411205,0.0076094150449207195
115−2020,quantum_computing,0.014749563377229326,0.014436233042737128,0.015025514431771204 115+2020,quantum_computing,0.014255991776269466,0.013671599966203846,0.014735782732982228
116−2020,reinforcement_learning,0.010335074967993138,0.01009190476797529,0.010581887485180323 116+2020,reinforcement_learning,0.009989227476272284,0.009571044089897765,0.010416211944123692
117−2020,robotics,0.03706604067375583,0.036544233210283716,0.037618898891096535 117+2020,robotics,0.03582568225983608,0.03507406990170952,0.036706054439969577
118−2020,scientific_computing,0.035949670461354104,0.035486140072591384,0.03641957490516316 118+2020,scientific_computing,0.03474666967076914,0.033850939603458474,0.03563936336497028
119−2020,software_engineering,0.05054807767605595,0.049996084509042964,0.051122015804014195 119+2020,software_engineering,0.0488565634945222,0.04796413383276536,0.04981011666713605
120−2020,speech_audio,0.02275173419826225,0.022346389826279685,0.023155166353730917 120+2020,speech_audio,0.03298557326724307,0.0321855549609936,0.03387364639366885
121−2020,sustainable_computing,0.001154614542679762,0.0010726624063696843,0.001243077320930029 121+2020,sustainable_computing,0.003347931393809671,0.0031103021939335902,0.003601402540344157
122−2021,ai_accelerators,0.0022344669660267994,0.0021092366316618066,0.002359697300391792 122+2021,ai_accelerators,0.002754393651148362,0.002549302690734593,0.0029638951698505995
123−2021,ai_agents,0.00875181136733635,0.008529035547524913,0.00900781795087393 123+2021,ai_agents,0.007192150382395403,0.00683195478883717,0.007540180186007944
124−2021,ai_safety,0.004461777912947028,0.004269415174338515,0.0046515018695099915 124+2021,ai_safety,0.003666644123813192,0.0034356043788022683,0.00392561740465108
125−2021,algorithms,0.05192228563250264,0.05128526575665957,0.0525297869295311 125+2021,algorithms,0.04266921106873708,0.041859726604193556,0.04348457627766522
126−2021,applied_cryptography,0.01954308818004544,0.0191733456178328,0.019934969676369034 126+2021,applied_cryptography,0.016060312914408705,0.015518439074017988,0.016593181864960463
127−2021,autonomous_scientific_discovery,0.00039537005563804856,0.00034970570871424225,0.0004481904216684258 127+2021,autonomous_scientific_discovery,0.000487366690875731,0.000401360804250602,0.0005821936940777963
128−2021,blockchain,0.019047533856915398,0.018727302001896345,0.019408912821797235 128+2021,blockchain,0.023479607048660218,0.022897082562710787,0.02407073212327216
129−2021,cloud_computing,0.013837951947331699,0.013552471510098932,0.014155634470544036 129+2021,cloud_computing,0.017057834180650586,0.016497362486143495,0.017505505846930104
130−2021,compilers,0.0012684043866397123,0.001184321162137503,0.00135785462547185 130+2021,compilers,0.0015635429132619606,0.0013981469774444048,0.0017180227173155575
131−2021,computational_biology,0.03438109379752044,0.033903429522156824,0.03482570620963558 131+2021,computational_biology,0.028254036395926992,0.02762266497693278,0.028929072341977044
132−2021,computational_complexity,0.0067356029840599674,0.006530717212014956,0.006953861566810384 132+2021,computational_complexity,0.005535250652027533,0.005230701602208807,0.0058330736004563455
133−2021,computer_architecture,0.02933967833694116,0.028900432939155948,0.029801420469792655 133+2021,computer_architecture,0.024111051976959244,0.02349713901785688,0.02477035693178042
134−2021,computer_graphics,0.016045583841708858,0.01569493890548688,0.01637757840313434 134+2021,computer_graphics,0.019779148644635435,0.019214266391839877,0.020317788075614608
135−2021,computer_networks,0.09239136268493837,0.09160674991502227,0.09308527291267868 135+2021,computer_networks,0.07592629074988312,0.07482026974974491,0.07702665153355111
136−2021,computer_vision,0.16654918868633378,0.16563004275721416,0.1674919942036245 136+2021,computer_vision,0.13686844480774374,0.13533566553854387,0.13833065514432816
137−2021,cybersecurity,0.05685278279693007,0.056236828452331966,0.05750049197631357 137+2021,cybersecurity,0.07008156592461473,0.0690958061471421,0.07105938669716812
138−2021,database_systems,0.025640016458843944,0.025257929763672467,0.026030198400629728 138+2021,database_systems,0.031606060695162794,0.03094403589606372,0.032227949413836804
139−2021,deep_learning_architectures,0.037969837379465805,0.037445569529670646,0.03844401309551496 139+2021,deep_learning_architectures,0.031203229704816034,0.03043472667770286,0.03187576633450262
140−2021,distributed_computing,0.018061792224985242,0.017730647440828667,0.01837343685707641 140+2021,distributed_computing,0.014842998826791494,0.014365188345540778,0.015302542245797471
141−2021,edge_computing,0.014151027783244181,0.013819122672057534,0.014447242249136806 141+2021,edge_computing,0.01744375803089155,0.01690335437659699,0.017962108893743767
142−2021,federated_learning,0.0011646421095944327,0.001075191870762295,0.0012434477700055458 142+2021,federated_learning,0.0014356367228963841,0.0012855674771312551,0.0015768848520845767
143−2021,formal_verification,0.008070200547435461,0.007818711200958907,0.00830652807842997 143+2021,formal_verification,0.00994801421963992,0.009553269252822022,0.010365694089557854
144−2021,foundation_models,0.00044546218938404565,0.00038821403653147756,0.0004964935506377801 144+2021,foundation_models,0.00036607633794285676,0.0002955074053273663,0.00044546638713528357
145−2021,generative_ai,0.003506449362219797,0.003341503121813335,0.003671977029178668 145+2021,generative_ai,0.004322347122698791,0.004073040315443128,0.004600432822786708
146−2021,hci,0.06159364545503337,0.06099276347567848,0.062190502173640805 146+2021,hci,0.07592555565683504,0.07497508034567014,0.07705939992884299
147−2021,information_retrieval,0.020650482136787305,0.02030162620534197,0.02101722811599907 147+2021,information_retrieval,0.02545553716189395,0.02480641324578865,0.026071251098964104
148−2021,llm,0.005096874608655205,0.004914306671198812,0.005278503318603861 148+2021,llm,0.006282840281922886,0.005931980370075244,0.006567321291529081
149−2021,ml_theory,0.03570137932268279,0.035164498989212305,0.036212274361772545 149+2021,ml_theory,0.02933903373489016,0.028752429482523895,0.030069311923503275
150−2021,multimodal_ai,0.0016959765282573305,0.0015848793316278155,0.0018051952698713706 150+2021,multimodal_ai,0.002090604628733905,0.0019162773123822011,0.0023111325431573126
151−2021,nlp,0.05294201835518901,0.052335992987101274,0.05346203731863964 151+2021,nlp,0.06526082571531903,0.06426811930854207,0.06620942653921133
152−2021,operating_systems,0.002366853319498363,0.002243411989910013,0.0025029071327620444 152+2021,operating_systems,0.002917584307821684,0.0027013566877295324,0.003149138617966262
153−2021,privacy_preserving_computation,0.0075871692577419185,0.007375977243859241,0.0077814104513659055 153+2021,privacy_preserving_computation,0.00935258885069672,0.008951117782488907,0.009758691005107425
154−2021,programming_languages,0.007116661001484874,0.006905468987602197,0.007313451526915577 154+2021,programming_languages,0.005848400290508772,0.005548261798978514,0.00616221151273328
155−2021,quantum_computing,0.016601964327244752,0.01626630230602716,0.016954666618959873 155+2021,quantum_computing,0.013643326972328156,0.013143243171720675,0.01413374429935341
156−2021,reinforcement_learning,0.011793119487629032,0.011537291804569118,0.012061559654364278 156+2021,reinforcement_learning,0.014537200118791036,0.014053692666417714,0.015000529266994614
157−2021,robotics,0.039476179401399004,0.03896443458504034,0.03996824516521459 157+2021,robotics,0.048661689596669146,0.047878705236508834,0.049510832331156475
158−2021,scientific_computing,0.036315007961071254,0.03584092169526092,0.03677657119344509 158+2021,scientific_computing,0.029843307565871686,0.02914496917019756,0.03047155483941157
159−2021,software_engineering,0.044136536844553376,0.04361235844499705,0.04464649265613539 159+2021,software_engineering,0.054406441767398915,0.0535198092874596,0.05529351530316707
160−2021,speech_audio,0.02287600407893089,0.02246359375279532,0.02326193713437215 160+2021,speech_audio,0.028198904417321143,0.02751284207554992,0.028875925114601003
161−2021,sustainable_computing,0.0012827164248528544,0.0011887489489596937,0.001383839919852586 161+2021,sustainable_computing,0.0015811851464158332,0.00143343144375215,0.0017421705239449208
162−2022,ai_accelerators,0.002535539484648477,0.0023954246091861114,0.002663164017821934 162+2022,ai_accelerators,0.0027899579837432427,0.002599314127148261,0.0030002356051780778
163−2022,ai_agents,0.008066976784022354,0.007835013284371192,0.008284665606771907 163+2022,ai_agents,0.00887642508409796,0.008546479665964214,0.00923778452597547
164−2022,ai_safety,0.005270924445919495,0.005080580551109589,0.005467379686777902 164+2022,ai_safety,0.005799814133692849,0.005513161167030982,0.006086467100354717
165−2022,algorithms,0.050321804747400674,0.04975072845460564,0.050886100568667436 165+2022,algorithms,0.03691409573423735,0.036232149635466435,0.03764970745690389
166−2022,applied_cryptography,0.01837686217621018,0.018004426934174114,0.01874207086306481 166+2022,applied_cryptography,0.020220814408565557,0.019649275514077406,0.020792156965405307
167−2022,autonomous_scientific_discovery,0.00042110296859749515,0.000370204823770192,0.0004791384518756033 167+2022,autonomous_scientific_discovery,0.0004633568502205526,0.00037696828492519535,0.000530111650676056
168−2022,blockchain,0.02100161839149372,0.02059884946104173,0.021372001648725183 168+2022,blockchain,0.02310894121650807,0.02253357373787615,0.023682443487480203
169−2022,cloud_computing,0.012674128754016984,0.012406210911919891,0.012960648284451487 169+2022,cloud_computing,0.013945863165748243,0.013480346601395306,0.014405587769473422
170−2022,compilers,0.001145542821354203,0.001053604980434771,0.0012419861071707055 170+2022,compilers,0.001260487702718622,0.0011348316077435569,0.001399985601905784
171−2022,computational_biology,0.034858760993731634,0.034417851910933155,0.03532184043409292 171+2022,computational_biology,0.03835652299113863,0.037619765965523105,0.03915178863597691
172−2022,computational_complexity,0.006318328863575129,0.0061228996151190245,0.006528032788932843 172+2022,computational_complexity,0.006952316129792274,0.00660479849212686,0.007294041806829932
173−2022,computer_architecture,0.027075493413128777,0.026614956649590585,0.027507391606133115 173+2022,computer_architecture,0.029792274767994346,0.029141906307674183,0.03044863152659066
174−2022,computer_graphics,0.014786780935455265,0.014468232598722774,0.015105017013630533 174+2022,computer_graphics,0.016270500922786948,0.015806260553148602,0.01675545491433134
175−2022,computer_networks,0.08768398720275215,0.0869405887947355,0.08839789948129392 175+2022,computer_networks,0.06432152253301744,0.06338633359075381,0.06524506210814278
176−2022,computer_vision,0.17612096361206425,0.1750984506622558,0.17715101537561137 176+2022,computer_vision,0.19379311247529418,0.19239499208104818,0.1954822052644668
177−2022,cybersecurity,0.056302894369175266,0.05570777416747408,0.05694926958262629 177+2022,cybersecurity,0.06195238157567507,0.06096372334716423,0.06289872904095603
178−2022,database_systems,0.02431155909805454,0.023886619810039738,0.02470964415014819 178+2022,database_systems,0.02675100459430097,0.026071578162017824,0.02739116349690441
179−2022,deep_learning_architectures,0.03973713182485685,0.03925085603453045,0.04019972057319965 179+2022,deep_learning_architectures,0.029149596198903123,0.028498311496223772,0.02981043600047121
180−2022,distributed_computing,0.01673527433252503,0.016396250756111792,0.017053242760508392 180+2022,distributed_computing,0.018414508043298996,0.017852884200054972,0.018901523580151573
181−2022,edge_computing,0.015263198277046497,0.014961556510769351,0.015591515845783529 181+2022,edge_computing,0.016794722444011048,0.016305743530674485,0.017344467859526955
182−2022,federated_learning,0.001737941912771018,0.001631729394949976,0.0018512917690428651 182+2022,federated_learning,0.0019123286954017722,0.0017100027487270775,0.002067533606460817
183−2022,formal_verification,0.0071284167777414965,0.006915813308638142,0.007356900824897891 183+2022,formal_verification,0.007843689053521643,0.007511878427728111,0.008219086637259648
184−2022,foundation_models,0.0006245171144454377,0.0005602810683881927,0.0006932586053997534 184+2022,foundation_models,0.0006871817693948874,0.0005929396981635885,0.0007913388918703123
185−2022,generative_ai,0.0038541627634347013,0.003694420207232622,0.004006767981185976 185+2022,generative_ai,0.004240893205408448,0.00400136127436223,0.004502120446602704
186−2022,hci,0.06209662885661622,0.06150141943818441,0.06272065527904316 186+2022,hci,0.06832746501917564,0.06732732103823348,0.06919724080158117
187−2022,information_retrieval,0.02006127405060016,0.019709760131898013,0.020421798859096448 187+2022,information_retrieval,0.022074241809447767,0.021508691213235773,0.02262222018612809
188−2022,llm,0.005827636845082284,0.005632118379895545,0.006025697987092123 188+2022,llm,0.004274925064464195,0.004020995039202084,0.004525680964410529
189−2022,ml_theory,0.035920440088288874,0.03539303538513971,0.03636661295819482 189+2022,ml_theory,0.039524731999109934,0.03872357622481969,0.04022398853388133
190−2022,multimodal_ai,0.002280379635032198,0.0021545394364714428,0.0024159890655974935 190+2022,multimodal_ai,0.0025091951465333316,0.0023324912629746463,0.002725166559771725
191−2022,nlp,0.052955482635747717,0.052373700335276475,0.053590794974599995 191+2022,nlp,0.038846058194478984,0.03814595086323119,0.039605230434953335
192−2022,operating_systems,0.0019324343855554546,0.0018075309626663669,0.0020493529110527037 192+2022,operating_systems,0.00212633673215618,0.0019299009149334413,0.0022971504862629093
193−2022,privacy_preserving_computation,0.007808248265180673,0.007590291792239215,0.008041148540503503 193+2022,privacy_preserving_computation,0.008591735493920076,0.008224289584942212,0.008937289755101507
194−2022,programming_languages,0.006537802020937383,0.006334298658358805,0.006740457824574927 194+2022,programming_languages,0.007193811437322478,0.00686150996740795,0.00753357373787615
195−2022,quantum_computing,0.016444427790654727,0.016103619879628786,0.016776492462078428 195+2022,quantum_computing,0.018094477676409378,0.017562108142776738,0.01861693215879789
196−2022,reinforcement_learning,0.012520675977324676,0.012244104112355983,0.012800058169308374 196+2022,reinforcement_learning,0.013777012788125499,0.01326064477283734,0.014222895587638581
197−2022,robotics,0.04122348255723699,0.04069250918486242,0.04169106744249535 197+2022,robotics,0.04535988690951452,0.044546950876320365,0.04622200552363251
198−2022,scientific_computing,0.03556535750036133,0.035086888173965484,0.03607135018815808 198+2022,scientific_computing,0.026089346719197895,0.025464403984345345,0.02674324925718923
199−2022,software_engineering,0.04233333868633717,0.04178612786898702,0.042844595161241394 199+2022,software_engineering,0.04658110708255343,0.04577916595766961,0.04736655584497179
200−2022,speech_audio,0.02265748091208048,0.022302041457230394,0.023059223850130168 200+2022,speech_audio,0.024930954593646512,0.024324173091270825,0.025494738151022916
201−2022,sustainable_computing,0.001480997728542038,0.0013775063210053654,0.0015765042386868724 201+2022,sustainable_computing,0.0010864016544719173,0.0009756999437165408,0.001199721200539274
202−2023,ai_accelerators,0.0025202296104125358,0.0023894109377019415,0.0026408428505231172 202+2023,ai_accelerators,0.0027909590845514565,0.0025953824957934885,0.002963932221061181
203−2023,ai_agents,0.007866457777091286,0.007650370547830774,0.00807773781416183 203+2023,ai_agents,0.00871149267729601,0.008433967665136108,0.009047228124396367
204−2023,ai_safety,0.006328156276301646,0.006121604624991725,0.006525329962519662 204+2023,ai_safety,0.007007942917119511,0.006693676386054164,0.007327358497152605
205−2023,algorithms,0.044968524712120116,0.04443327144401749,0.04549408478941346 205+2023,algorithms,0.033199438666390496,0.03251336425477843,0.03380088831276428
206−2023,applied_cryptography,0.01768258661614649,0.01736578477018469,0.01799107438381253 206+2023,applied_cryptography,0.013054730026926325,0.01263995522363745,0.013464501234608123
207−2023,autonomous_scientific_discovery,0.0006304514347498526,0.0005657513562586491,0.0006934965782248378 207+2023,autonomous_scientific_discovery,0.0006981761312198765,0.0006090714024729398,0.0007977535019351114
208−2023,blockchain,0.022615869093064088,0.022195712615017984,0.022956431077473024 208+2023,blockchain,0.025045323267185022,0.024498825900472664,0.025574540774649688
209−2023,cloud_computing,0.01204635078948281,0.011778408929714122,0.012319099841441463 209+2023,cloud_computing,0.013340400427283792,0.01292306564484711,0.013749095280096626
210−2023,compilers,0.001090680982117245,0.001000014185157282,0.0011671626217953366 210+2023,compilers,0.0012078447070103865,0.0010803402910463564,0.0013371818353188683
211−2023,computational_biology,0.03243042180353242,0.03197551169014573,0.032867836889604804 211+2023,computational_biology,0.03591418018995045,0.03518057162007116,0.03662850164420479
212−2023,computational_complexity,0.005514873925474336,0.005303593888403791,0.00568434715177803 212+2023,computational_complexity,0.004071530471897246,0.0038609140056459173,0.004288634158034494
213−2023,computer_architecture,0.025799648838550845,0.025440212714314084,0.026166216944643213 213+2023,computer_architecture,0.0285711127298454,0.027938477882943787,0.029237958207176784
214−2023,computer_graphics,0.014653267472173449,0.01434028773803482,0.014952140855459551 214+2023,computer_graphics,0.010818239153251988,0.010441224042393255,0.011221523141047868
215−2023,computer_networks,0.08177112721564277,0.08109721403510985,0.08245847689237878 215+2023,computer_networks,0.060370126439697365,0.059414439678745885,0.06123074488411439
216−2023,computer_vision,0.18009002846488228,0.17915365047142007,0.18112672704289906 216+2023,computer_vision,0.19943575732328578,0.19807204479498058,0.20088141827499295
217−2023,cybersecurity,0.057498746977773435,0.05688472668354175,0.05807135449338499 217+2023,cybersecurity,0.04245027240505386,0.04168277320213828,0.04324686228010361
218−2023,database_systems,0.022833374838052787,0.0224598323629635,0.02320849344172895 218+2023,database_systems,0.02528619403245588,0.024732889448464128,0.025859658452236606
219−2023,deep_learning_architectures,0.04035992472409869,0.03982313473062386,0.04082338533506918 219+2023,deep_learning_architectures,0.04469549048036845,0.04391676227800908,0.045337899793107136
220−2023,distributed_computing,0.01544290789419764,0.015133080417232759,0.01573449168276945 220+2023,distributed_computing,0.011401216222820584,0.010990775929679699,0.011794376656713776
221−2023,edge_computing,0.01437429271229664,0.014125185589141104,0.014658823325442183 221+2023,edge_computing,0.010612277194542124,0.010263014584899382,0.010982485088121463
222−2023,federated_learning,0.0023594644945513234,0.002249135493470099,0.0024753493489012806 222+2023,federated_learning,0.002612924171090388,0.0024294783926123655,0.002799686286191705
223−2023,formal_verification,0.006911323853445259,0.006714228973656587,0.007113895780073321 223+2023,formal_verification,0.005102503892331932,0.004824222522696542,0.005374298042081402
224−2023,foundation_models,0.0008999694231054147,0.0008313684263616964,0.0009780271913703807 224+2023,foundation_models,0.0009966464273163738,0.000884850974304791,0.0011189145222962546
225−2023,generative_ai,0.0057465648277449065,0.005570787087093714,0.0059168655215567105 225+2023,generative_ai,0.006363875436069174,0.006060168818988528,0.006671072933805921
226−2023,hci,0.06397033095548067,0.06340178197098031,0.06462765538263673 226+2023,hci,0.07084218659455283,0.06984283473472798,0.07180043333465211
227−2023,information_retrieval,0.01980878407984037,0.01948063410805307,0.020147651726018415 227+2023,information_retrieval,0.02193669404292852,0.021444479870418507,0.022396530297353213
228−2023,llm,0.016374399889040547,0.016070876926423165,0.016689704412844817 228+2023,llm,0.018133379568108243,0.017649892597238478,0.01864863355294851
229−2023,ml_theory,0.036082311739320944,0.0356795714821598,0.0365213029539802 229+2023,ml_theory,0.03995836543004159,0.039268218324330736,0.04065680337731067
230−2023,multimodal_ai,0.0030450804298417884,0.0029047261791806025,0.003182282423329225 230+2023,multimodal_ai,0.003372190713792004,0.0031697196357382395,0.0035854835218796757
231−2023,nlp,0.05353320745319686,0.052956620212903455,0.05405581228938982 231+2023,nlp,0.05928388074220777,0.05844510939256349,0.06009402687047203
232−2023,operating_systems,0.0018487988324039429,0.0017376423638146095,0.0019496710619639193 232+2023,operating_systems,0.002047401504802288,0.0018920573156058654,0.0022271818585914063
233−2023,privacy_preserving_computation,0.007896404220241904,0.0076685354297945044,0.008110836514486198 233+2023,privacy_preserving_computation,0.008744656043528955,0.008376106318261262,0.009052202629331308
234−2023,programming_languages,0.006109074402726072,0.005919702553013085,0.006301362090324777 234+2023,programming_languages,0.006765326711520604,0.006489547139688753,0.0070692078726340545
235−2023,quantum_computing,0.016618699820006116,0.016293189863601832,0.016949686823249786 235+2023,quantum_computing,0.018403922818955948,0.01791685769541368,0.01886140272893777
236−2023,reinforcement_learning,0.013274154958658147,0.012999002310604508,0.013544500414521817 236+2023,reinforcement_learning,0.014700098442834502,0.014228829554261085,0.01516970916309627
237−2023,robotics,0.040599496269303634,0.04014001538301501,0.04108135818152588 237+2023,robotics,0.044960797410232,0.04425502861358511,0.045807510513368906
238−2023,scientific_computing,0.032189274129740604,0.03178097801931073,0.032598397707678584 238+2023,scientific_computing,0.023764751879839233,0.023231636222308588,0.02435072528863765
239−2023,software_engineering,0.041365494762524706,0.040874254885210556,0.04187564818288136 239+2023,software_engineering,0.04580908140966415,0.04505496391593028,0.046598107709959015
240−2023,speech_audio,0.02315017668401459,0.022767177437404054,0.02356816598525374 240+2023,speech_audio,0.025637027538393867,0.025080144801729612,0.026157168756152677
241−2023,sustainable_computing,0.001699066616650853,0.0015910624052352688,0.001802342442305813 241+2023,sustainable_computing,0.0018815846736375673,0.001722662331768643,0.0020351834225059403
242−2024,ai_accelerators,0.002325521653716305,0.002229526283126853,0.002434395276486948 242+2024,ai_accelerators,0.002366081397785513,0.002207815313590616,0.0025120356049396077
243−2024,ai_agents,0.007975728043763073,0.007740404181071606,0.00819283306675464 243+2024,ai_agents,0.008114833817172524,0.007832829929645763,0.00840247778244982
244−2024,ai_safety,0.008405679140205971,0.008204798686891494,0.00860307807127724 244+2024,ai_safety,0.008552283750018916,0.008264639784741618,0.008864551469377818
245−2024,algorithms,0.04050328615138335,0.04007654616768904,0.04099407938411533 245+2024,algorithms,0.02747313970289171,0.02695957164755842,0.027972309510959407
246−2024,applied_cryptography,0.016800541900627214,0.016508364441692278,0.017096775501981423 246+2024,applied_cryptography,0.01709356247028069,0.01670838642878072,0.01751044675376842
247−2024,autonomous_scientific_discovery,0.000616533647729439,0.0005665078912250767,0.0006746716890723467 247+2024,autonomous_scientific_discovery,0.0006272866961570896,0.0005460558202621857,0.0007180781916535104
248−2024,blockchain,0.020789081946245295,0.020467598858331115,0.021094002452614117 248+2024,blockchain,0.021151667193226774,0.02068395342854824,0.021608513491020127
249−2024,cloud_computing,0.011639776695839344,0.01140715692809406,0.011890682905658184 249+2024,cloud_computing,0.011842787647404351,0.011470886423048936,0.012178509836433161
250−2024,compilers,0.0010383724593337922,0.0009626578008407031,0.001112092847804072 250+2024,compilers,0.0010564828566856244,0.0009491838165534908,0.001170728821851954
251−2024,computational_biology,0.03180286066204357,0.03137050968133593,0.03217196962219737 251+2024,computational_biology,0.03235753874343653,0.03176024075336766,0.032918932823922734
252−2024,computational_complexity,0.0051702295370995065,0.0050180903951899565,0.005342716993479075 252+2024,computational_complexity,0.005260404223913839,0.0050306054462975185,0.0054929542589694435
253−2024,computer_architecture,0.023199782590766327,0.022850886740335228,0.023551382536143608 253+2024,computer_architecture,0.02360441320034978,0.02309811305008527,0.024094480931722503
254−2024,computer_graphics,0.014346575736967276,0.014061935942694821,0.014593425804366844 254+2024,computer_graphics,0.014596796344129116,0.014219805293511023,0.014987818807687564
255−2024,computer_networks,0.07347296378270433,0.07286582686491287,0.07408848339482896 255+2024,computer_networks,0.07475441588282579,0.07388329899611203,0.07557497841409268
256−2024,computer_vision,0.18025091297005622,0.1793576490057719,0.18106481174767008 256+2024,computer_vision,0.18339469401661118,0.18207175187693078,0.18470711259658634
257−2024,cybersecurity,0.06158305830434318,0.06101100701847843,0.062084803121607204 257+2024,cybersecurity,0.06265713692228753,0.06193823335340223,0.06346593782800149
258−2024,database_systems,0.02031316123571731,0.01999742434956376,0.020655432053530265 258+2024,database_systems,0.02066744588391253,0.02018308701172632,0.021099668427624275
259−2024,deep_learning_architectures,0.03918909600753902,0.038687790605703744,0.03962181880130176 259+2024,deep_learning_architectures,0.039872598438844824,0.0392177303868589,0.040469690084605756
260−2024,distributed_computing,0.014333055262236367,0.01406663430766381,0.01457983772726228 260+2024,distributed_computing,0.014583040056932688,0.014178261306177813,0.014947581667638016
261−2024,edge_computing,0.013804404700257835,0.01351827765376498,0.014039018738025929 261+2024,edge_computing,0.014045169227552378,0.013651464287990625,0.014440043451525825
262−2024,federated_learning,0.0027987382692981114,0.0026769863943462784,0.002915149556731236 262+2024,federated_learning,0.0028475514496604723,0.002675253952525219,0.003027827593369653
263−2024,formal_verification,0.005723216953593674,0.005543327037298934,0.005893642537576779 263+2024,formal_verification,0.0038820242468318128,0.003682214175303705,0.0040801147824603675
264−2024,foundation_models,0.0014142416568530553,0.0013331188084676029,0.0014994206476577805 264+2024,foundation_models,0.0014389076407463062,0.0013095985410998886,0.0015753012282988838
265−2024,generative_ai,0.009663083290180485,0.009420999190123564,0.00988218258319486 265+2024,generative_ai,0.009831618459286665,0.00944362237891143,0.010120569271847622
266−2024,hci,0.06499833022137073,0.06439812254687886,0.06555648921944947 266+2024,hci,0.06613197506810509,0.06536513083834024,0.06690535353428824
267−2024,information_retrieval,0.01912606355434352,0.01885024586983298,0.01942834756813981 267+2024,information_retrieval,0.019459643868066204,0.01899021056748812,0.01991373890842027
268−2024,llm,0.03362947679818934,0.0332200430221506,0.0341027610161448 268+2024,llm,0.034216013143673875,0.03362284203976393,0.03480553883147677
269−2024,ml_theory,0.03611318800625728,0.03570750616195636,0.036529686230342924 269+2024,ml_theory,0.03674304310165759,0.03609333365737032,0.03742274125203307
270−2024,multimodal_ai,0.004607777788293702,0.004451582503964879,0.0047653251200956166 270+2024,multimodal_ai,0.0046881426765424585,0.004473407033406227,0.00489448698448887
271−2024,nlp,0.052501355427591774,0.05198176358368295,0.052959767123343236 271+2024,nlp,0.05341703881244724,0.052638226612821505,0.05413663019569236
272−2024,operating_systems,0.0017035798160945027,0.001608868890604487,0.0018076874715225002 272+2024,operating_systems,0.0017332921867498526,0.0016012318296641495,0.0018627388492682344
273−2024,privacy_preserving_computation,0.009039789405085591,0.008820047889521497,0.00924137968332344 273+2024,privacy_preserving_computation,0.009197453619531362,0.008862969496350228,0.009505869578475263
274−2024,programming_languages,0.005199974581507506,0.0050375260776156375,0.005375808355382974 274+2024,programming_languages,0.003527112037163986,0.0033314976332307882,0.0037528527100573594
275−2024,quantum_computing,0.017200747952662115,0.016901742653988066,0.01751141466079157 275+2024,quantum_computing,0.017500748571294942,0.017088059955402118,0.01788344848109955
276−2024,reinforcement_learning,0.013374453603814937,0.013108674871792098,0.013619174196444386 276+2024,reinforcement_learning,0.013607719294705986,0.013212432382116645,0.013968134019252385
277−2024,robotics,0.04054249552810298,0.04004081831321261,0.04097863224173527 277+2024,robotics,0.041249602787207205,0.04064693984513172,0.04187991586654751
278−2024,scientific_computing,0.029592263043539985,0.029195775122056088,0.029967557620883184 278+2024,scientific_computing,0.030108385786820655,0.02956081677496686,0.030610490269490254
279−2024,software_engineering,0.04128206549588369,0.04084998492467068,0.04173250011154392 279+2024,software_engineering,0.04200207169685179,0.04136371118950157,0.04267475414075709
280−2024,speech_audio,0.022210083840463805,0.02187920402261164,0.022523116631671173 280+2024,speech_audio,0.022597452977571292,0.02208393077652866,0.023041781054015897
281−2024,sustainable_computing,0.0017184523382985024,0.0016190430478394957,0.0018076874715225002 281+2024,sustainable_computing,0.0017484241026659228,0.0016313580986243256,0.0018833044986268934
282−2025,ai_accelerators,0.0021347609874109584,0.002043451234263698,0.002224950719053806 282+2025,ai_accelerators,0.002218169672523209,0.0020761881929598136,0.0023531684563703713
283−2025,ai_agents,0.00987746964741723,0.009682249899198065,0.01009646185206756 283+2025,ai_agents,0.010263398920242486,0.00995831330657409,0.010557835922943626
284−2025,ai_safety,0.01219143407553425,0.011944973343488196,0.012440890865104609 284+2025,ai_safety,0.01266777381186523,0.012339296167547456,0.013030059341275761
285−2025,algorithms,0.03785112674163344,0.03743932283499843,0.03824159423860938 285+2025,algorithms,0.02622002241445325,0.025748515886214646,0.026684429868713683
286−2025,applied_cryptography,0.01636575422248107,0.016069620536714305,0.016610506921732895 286+2025,applied_cryptography,0.01700519163426502,0.016601300876211932,0.017405765611852895
287−2025,autonomous_scientific_discovery,0.000855696429371444,0.000796335289637561,0.0009111654943774921 287+2025,autonomous_scientific_discovery,0.0008891299212002788,0.0007948053399001707,0.0009845600959887905
288−2025,blockchain,0.02017382733748488,0.01991328233502083,0.020477353165180772 288+2025,blockchain,0.020962052540130133,0.020516032851255778,0.021342667645795777
289−2025,cloud_computing,0.012291115989426997,0.012060335558442723,0.012523772456431163 289+2025,cloud_computing,0.00851423364332623,0.00823946456689248,0.008794763443725873
290−2025,compilers,0.0008590564938846826,0.0008002273643653959,0.0009167656018995565 290+2025,compilers,0.0008926212690583952,0.0007924777746614266,0.0009764718067841543
291−2025,computational_biology,0.03153308543524036,0.0311824627032839,0.03189544039245553 291+2025,computational_biology,0.032765135865801896,0.032314461046450055,0.0333916000498099
292−2025,computational_complexity,0.004714170512073831,0.004580299941758882,0.004862013350656332 292+2025,computational_complexity,0.004898361044937139,0.0047062205344788055,0.005120759903499145
293−2025,computer_architecture,0.02263451458267999,0.022314076430267463,0.022954840732942072 293+2025,computer_architecture,0.015679255303260413,0.015315262892674803,0.0160718379735286
294−2025,computer_graphics,0.013933067514896286,0.0136916748801577,0.014181348281887012 294+2025,computer_graphics,0.014477455784988833,0.014120116331710631,0.01484998260144984
295−2025,computer_networks,0.06670288069530934,0.06617697459791229,0.06721394650777295 295+2025,computer_networks,0.06930907389670499,0.06849535708924001,0.07009253235606629
296−2025,computer_vision,0.16885108194077325,0.16800818175708973,0.16958203597509072 296+2025,computer_vision,0.17544837634867858,0.17428651397062844,0.17643322739032222
297−2025,cybersecurity,0.06613054970655437,0.06558070314950047,0.06666502396846019 297+2025,cybersecurity,0.06871438097820584,0.06803339357848026,0.06940590061063674
298−2025,database_systems,0.01868307871511133,0.018413153532547823,0.018980752430446664 298+2025,database_systems,0.01941305787374588,0.0189342777041362,0.01982509511013457
299−2025,deep_learning_architectures,0.0382386541821603,0.03786392298732136,0.038624221585054436 299+2025,deep_learning_architectures,0.039732702407982615,0.03915296409614241,0.04029259822616253
300−2025,distributed_computing,0.014056269880381703,0.01381263720263429,0.014313146812418798 300+2025,distributed_computing,0.009736981248746508,0.009414768634196356,0.010062064527091113
301−2025,edge_computing,0.01294072846198647,0.012697039783163836,0.013170388871466331 301+2025,edge_computing,0.01344634438422516,0.013122812816039718,0.013771097924160942
302−2025,federated_learning,0.002849334707226379,0.0027412246315129253,0.0029546727297164103 302+2025,federated_learning,0.002960662983682604,0.002803494140936403,0.003127142087383782
303−2025,formal_verification,0.0051778594149007665,0.005019880381703329,0.00532858630885713 303+2025,formal_verification,0.005380167049357185,0.005183487786683301,0.005612981762362572
304−2025,foundation_models,0.0020765198691814883,0.0019823820617355854,0.0021633495363110973 304+2025,foundation_models,0.00215765297631586,0.0020214322207183564,0.002289218601435875
305−2025,generative_ai,0.01427243403073339,0.014036641503516868,0.014530094977823574 305+2025,generative_ai,0.014830081918658578,0.014443647899896075,0.015152798839010459
306−2025,hci,0.06512701043860042,0.06458724407508624,0.06561657183817929 306+2025,hci,0.06767163175124845,0.06691534761604949,0.06846114188023047
307−2025,information_retrieval,0.019464853725191522,0.019179780251780836,0.01973494691098069 307+2025,information_retrieval,0.0202253781420676,0.019800248351210973,0.02066080740910567
308−2025,llm,0.05057793109627705,0.05014526678912235,0.05100300725773935 308+2025,llm,0.052554095525604964,0.05185338201048102,0.053167467155145025
309−2025,ml_theory,0.03673446530173379,0.03632224138703463,0.037132716948165403 309+2025,ml_theory,0.038169742350165896,0.03751558013981684,0.03875751076207977
310−2025,multimodal_ai,0.007559025133282559,0.007395978002777653,0.007723780296581694 310+2025,multimodal_ai,0.007854368898142254,0.007602933663226913,0.008127857813694695
311−2025,nlp,0.04806460284037453,0.04766080708749608,0.04852733972492272 311+2025,nlp,0.049942567327733986,0.049287357713027495,0.05058974684236681
312−2025,operating_systems,0.001526589310514762,0.001447067783701447,0.0016083508803369025 312+2025,operating_systems,0.0015862357102041625,0.0014755599831018764,0.0017073272917498285
313−2025,privacy_preserving_computation,0.009875229604408405,0.009669509654585368,0.010055553066618879 313+2025,privacy_preserving_computation,0.010261071355003742,0.009930266145447224,0.01056260743168305
314−2025,programming_languages,0.004696250168003226,0.004548351328345504,0.0048334808028314144 314+2025,programming_languages,0.0048797405230271845,0.004658563636215519,0.005091607148883875
315−2025,quantum_computing,0.01956789570359751,0.01930452264683482,0.0198592132968953 315+2025,quantum_computing,0.020332446143049833,0.01989468931077302,0.02074215581419978
316−2025,reinforcement_learning,0.014795484073294207,0.014556863491779043,0.015040236772546033 316+2025,reinforcement_learning,0.015373568401905345,0.015026528424808587,0.015750633970581903
317−2025,robotics,0.03748039962367278,0.037061287576721474,0.037860310917969625 317+2025,robotics,0.038944821574667714,0.038327784029876626,0.03951874097341106
318−2025,scientific_computing,0.02597441870883921,0.025647204426324986,0.026275872496751937 318+2025,scientific_computing,0.017992855150572133,0.01758428986566457,0.01838433222735192
319−2025,software_engineering,0.04074750235204516,0.04033581044756059,0.04116639039469558 319+2025,software_engineering,0.042339575475376104,0.041742205856852406,0.042954110887535535
320−2025,speech_audio,0.020578155100577932,0.020276869315890865,0.020856508444962143 320+2025,speech_audio,0.02138217806572346,0.02093255065072905,0.021811730230533698
321−2025,sustainable_computing,0.0018357152457327182,0.0017461135253796873,0.0019159367859862908 321+2025,sustainable_computing,0.00190743971315086,0.001785126159854853,0.0020262037294577822
modified results/bibliometrics/diversity_yearly.csv +8 −8
@@ -1,9 +1,9 @@
1 year,n,rao_stirling,rs_ci_low,rs_ci_high,leinster_cobbold_q2,similarity_source 1 year,n,rao_stirling,rs_ci_low,rs_ci_high,leinster_cobbold_q2,similarity_source
2−2018,499561,0.8966979269920947,0.8964150650642733,0.8969729208682462,9.680347846683333,co-classification Jaccard 2+2018,169727,0.8992868732205007,0.8988175897191749,0.8997310848387939,9.929192270928045,co-classification Jaccard
3−2019,539741,0.8957695076396416,0.8955171860453549,0.896020741688698,9.594121426028362,co-classification Jaccard 3+2019,180248,0.896100206436324,0.8956356536840846,0.8965579428776569,9.624658199028467,co-classification Jaccard
4−2020,549101,0.896412402094723,0.8961367144172547,0.8966751994754848,9.653665305709902,co-classification Jaccard 4+2020,189370,0.900723632618864,0.9002795125878724,0.9011485807064236,10.0728907229337,co-classification Jaccard
5−2021,558970,0.8939562806724264,0.8936855340180797,0.8942329598613235,9.43007286373045,co-classification Jaccard 5+2021,226728,0.9022200511081365,0.9018514633225768,0.9026284736924316,10.22704564006183,co-classification Jaccard
6−2022,560433,0.8914726303600234,0.891161796756561,0.8917985624313133,9.214265519539925,co-classification Jaccard 6+2022,254663,0.890309885996106,0.8897676301389984,0.8907889622855558,9.116591855894152,co-classification Jaccard
7−2023,634466,0.8905938872892762,0.8902601885615422,0.8908748156052915,9.140257113823784,co-classification Jaccard 7+2023,286460,0.8855095802901976,0.8849779485027648,0.8860485908545297,8.73435526338963,co-classification Jaccard
8−2024,739619,0.8900517958692582,0.889784104970656,0.890332496245109,9.095191757846983,co-classification Jaccard 8+2024,363470,0.8883809706057654,0.8879340384933714,0.8887799732899944,8.959045831405975,co-classification Jaccard
9−2025,892840,0.893113751756877,0.892858747038931,0.893398088972461,9.355740485206343,co-classification Jaccard 9+2025,429633,0.889452772558058,0.8891043925701015,0.8898464283041351,9.04590755589223,co-classification Jaccard
modified tables/make_tables.py +58 −10
@@ -92,6 +92,7 @@ def main():
92 if os.path.exists(p): 92 if os.path.exists(p):
93 d = pd.read_csv(p)[["concept", "total_works", "detection_month", "peak_growth_month", "lead_months"]]; write("tab_known_set", d, "Preregistered known-emergence set: detection month, peak-growth month and lead time.", "tab:known_set", [p]) 93 d = pd.read_csv(p)[["concept", "total_works", "detection_month", "peak_growth_month", "lead_months"]]; write("tab_known_set", d, "Preregistered known-emergence set: detection month, peak-growth month and lead time.", "tab:known_set", [p])
94 tables_part08() # Part VIII tables (appended 2026-11-03) 94 tables_part08() # Part VIII tables (appended 2026-11-03)
95+ tables_part05() # Part V tables (Agent E, appended 2026-11-03)
95 tables_part07b() # Part VII-b security tables (Agent H, appended 2026-11-03) 96 tables_part07b() # Part VII-b security tables (Agent H, appended 2026-11-03)
96 tables_part06() # Part VI systems table (appended by Agent F, 2026-11-03) 97 tables_part06() # Part VI systems table (appended by Agent F, 2026-11-03)
97 tables_part09() # Part IX results tables (appended by the Part IX writer, 2026-11-03) 98 tables_part09() # Part IX results tables (appended by the Part IX writer, 2026-11-03)
@@ -307,8 +308,8 @@ def _p9_names():
307 return {c["id"]: c["name"] for c in o["categories"]}, {c["id"]: c["area"] for c in o["categories"]}, {a["id"]: a["name"] for a in o["areas"]} 308 return {c["id"]: c["name"] for c in o["categories"]}, {c["id"]: c["area"] for c in o["categories"]}, {a["id"]: a["name"] for a in o["areas"]}
308 309
309 310
310−_P9_AREA_SHORT = {"ai_ml": "AI \\& ML", "perception_interaction": "Perception \\& interaction", "software_pl": "Software \\& PL", "theory": "Theory", "data_distributed": "Data \\& distributed", 311+_P9_AREA_SHORT = {"ai_ml": "AI & ML", "perception_interaction": "Perception & interaction", "software_pl": "Software & PL", "theory": "Theory", "data_distributed": "Data & distributed",
311− "security_privacy": "Security \\& privacy", "hardware_emerging": "Hardware \\& emerging", "computational_science": "Computational science"} 312+ "security_privacy": "Security & privacy", "hardware_emerging": "Hardware & emerging", "computational_science": "Computational science"}
312 _P9_EVENTS = [("gpt3", "GPT-3 (2020-06)"), ("dalle2", "DALL$\\cdot$E 2 (2022-05)"), ("chatgpt", "ChatGPT (2022-12)"), ("gpt4", "GPT-4 (2023-04)"), ("o1", "o1 (2024-10)")] 313 _P9_EVENTS = [("gpt3", "GPT-3 (2020-06)"), ("dalle2", "DALL$\\cdot$E 2 (2022-05)"), ("chatgpt", "ChatGPT (2022-12)"), ("gpt4", "GPT-4 (2023-04)"), ("o1", "o1 (2024-10)")]
313 314
314 315
@@ -328,15 +329,16 @@ def tables_part09():
328 s = m[mask].groupby("period").n.sum().reindex(["baseline", "post"]).fillna(0) / tot * 100 329 s = m[mask].groupby("period").n.sum().reindex(["baseline", "post"]).fillna(0) / tot * 100
329 rows.append({"Indicator": label, "Baseline": float(s["baseline"]), "Post": float(s["post"])}) 330 rows.append({"Indicator": label, "Baseline": float(s["baseline"]), "Post": float(s["post"])})
330 rows.append({"Indicator": "Eligible works (count)", "Baseline": float(tot["baseline"]), "Post": float(tot["post"])}) 331 rows.append({"Indicator": "Eligible works (count)", "Baseline": float(tot["baseline"]), "Post": float(tot["post"])})
331− share(m.date_precision.eq("month"), "Month-precision first date (\\%)"); share(m.in_dblp.astype(bool), "DBLP-indexed (\\%)"); share(m.cs_primary.astype(bool), "OpenAlex primary field CS (\\%)") 332+ share(m.date_precision.eq("month"), "Month-precision first date (%)"); share(m.in_dblp.astype(bool), "DBLP-indexed (%)"); share(m.cs_primary.astype(bool), "OpenAlex primary field CS (%)")
332− share(m.has_arxiv.astype(bool), "arXiv-linked (\\%)") 333+ share(m.has_arxiv.astype(bool), "arXiv-linked (%)")
333− for c, lab in [("n_abstract", "With abstract (\\%)"), ("n_with_refs", "With at least one reference (\\%)"), ("n_oa", "Open access (\\%)")]: 334+ for c, lab in [("n_abstract", "With abstract (%)"), ("n_with_refs", "With at least one reference (%)"), ("n_oa", "Open access (%)")]:
334 s = m.groupby("period")[c].sum() / tot * 100; rows.append({"Indicator": lab, "Baseline": float(s["baseline"]), "Post": float(s["post"])}) 335 s = m.groupby("period")[c].sum() / tot * 100; rows.append({"Indicator": lab, "Baseline": float(s["baseline"]), "Post": float(s["post"])})
335− for ty, lab in [("article", "Type: article (\\%)"), ("conference-paper", "Type: conference paper (\\%)"), ("preprint", "Type: preprint (\\%)"), ("book-chapter", "Type: book chapter (\\%)"), ("dissertation", "Type: dissertation (\\%)")]: 336+ for ty, lab in [("article", "Type: article (%)"), ("conference-paper", "Type: conference paper (%)"), ("preprint", "Type: preprint (%)"), ("book-chapter", "Type: book chapter (%)"), ("dissertation", "Type: dissertation (%)")]:
336 share(m.type.eq(ty), lab) 337 share(m.type.eq(ty), lab)
337 t = pd.DataFrame(rows) 338 t = pd.DataFrame(rows)
339+ for c in ("Baseline", "Post"): t[c] = [f"{v:,.0f}" if i == 0 else f"{v:.2f}" for i, v in enumerate(t[c])]
338 write("tab_p9_periods", t, "The eligible corpus by protocol period: completeness, source membership and document type.", "tab:p9-periods", [p], fmt={"Baseline": "{:,.2f}", "Post": "{:,.2f}"}, 340 write("tab_p9_periods", t, "The eligible corpus by protocol period: completeness, source membership and document type.", "tab:p9-periods", [p], fmt={"Baseline": "{:,.2f}", "Post": "{:,.2f}"},
339− note="Baseline: first date 2018-01-01 to 2022-11-29; post: 2022-11-30 to 2026-11-02 (right-truncated, see Part III). Percentages are over all eligible works of the period. The count row is an integer written with two decimals by the table generator.") 341+ note="Baseline: first date 2018-01-01 to 2022-11-29; post: 2022-11-30 to 2026-11-02 (right-truncated, see Part III). Percentages are over all eligible works of the period. The first row is a count.")
340 # --- classification coverage by year --- 342 # --- classification coverage by year ---
341 p = f"{G}/category_counts_yearly.csv"; q = f"{R}/data/metadata/prisma_counts.json" 343 p = f"{G}/category_counts_yearly.csv"; q = f"{R}/data/metadata/prisma_counts.json"
342 if os.path.exists(p) and os.path.exists(q): 344 if os.path.exists(p) and os.path.exists(q):
@@ -376,7 +378,7 @@ def tables_part09():
376 z = s[(s.category == k) & (s.year == yv)] 378 z = s[(s.category == k) & (s.year == yv)]
377 r[str(yv)] = f"{100 * float(z.share.iloc[0]):.2f} [{100 * float(z.ci_low.iloc[0]):.2f}, {100 * float(z.ci_high.iloc[0]):.2f}]" if len(z) else "" 379 r[str(yv)] = f"{100 * float(z.share.iloc[0]):.2f} [{100 * float(z.ci_low.iloc[0]):.2f}, {100 * float(z.ci_high.iloc[0]):.2f}]" if len(z) else ""
378 rows.append(r) 380 rows.append(r)
379− r = {"Category": "AI \\& ML area (sum)"} 381+ r = {"Category": "AI & ML area (sum)"}
380 for yv in (2018, 2022, 2025): 382 for yv in (2018, 2022, 2025):
381 z = s[(s.category.isin(ai)) & (s.year == yv)]; r[str(yv)] = f"{100 * float(z.share.sum()):.2f}" if len(z) else "" 383 z = s[(s.category.isin(ai)) & (s.year == yv)]; r[str(yv)] = f"{100 * float(z.share.sum()):.2f}" if len(z) else ""
382 rows.append(r) 384 rows.append(r)
@@ -388,9 +390,9 @@ def tables_part09():
388 J = json.load(open(p)); S = J["series"]; T = S["total"]; rows = [] 390 J = json.load(open(p)); S = J["series"]; T = S["total"]; rows = []
389 for key, lab in _P9_EVENTS: 391 for key, lab in _P9_EVENTS:
390 e = T.get("its_" + key) 392 e = T.get("its_" + key)
391− if e: rows.append({"Event (month)": lab, "Pre-slope (SE)": f"{e['pre_slope']:.4f} ({e['pre_slope_se']:.4f})", "Level change (SE)": f"{e['level_change']:+.3f} ({e['level_se']:.3f})", "HAC p": e["level_p"], "Slope change (SE)": f"{e['slope_change']:+.4f} ({e['slope_se']:.4f})", "HAC p ": e["slope_p"], "R2": e["r2"]}) 393+ if e: rows.append({"Event (month)": lab, "Pre-slope (SE)": f"{e['pre_slope']:.4f} ({e['pre_slope_se']:.4f})", "Level change (SE)": f"{e['level_change']:+.3f} ({e['level_se']:.3f})", "HAC p (level)": e["level_p"], "Slope change (SE)": f"{e['slope_change']:+.4f} ({e['slope_se']:.4f})", "HAC p (slope)": e["slope_p"], "$R^2$": e["r2"]})
392 write("tab_p9_its_total_events", pd.DataFrame(rows), "Interrupted time series of the total monthly series at the reference date and at four alternative event dates (log counts, month-of-year dummies, Newey--West standard errors).", "tab:p9-its-total-events", [p], 394 write("tab_p9_its_total_events", pd.DataFrame(rows), "Interrupted time series of the total monthly series at the reference date and at four alternative event dates (log counts, month-of-year dummies, Newey--West standard errors).", "tab:p9-its-total-events", [p],
393− fmt={"HAC p": "{:.3f}", "HAC p ": "{:.3g}", "R2": "{:.3f}"}, note="Equation~\\eqref{eq:its}; each event is coded as the first full month after the public release. Slopes are monthly log growth rates. A HAC $p$ printed as 0.000 is below 0.0005.") 395+ fmt={"HAC p (level)": "{:.3f}", "HAC p (slope)": "{:.3g}", "$R^2$": "{:.3f}"}, note="Equation~\\eqref{eq:its}; each event is coded as the first full month after the public release. Slopes are monthly log growth rates. A HAC $p$ printed as 0.000 is below 0.0005.")
394 rows = [("Observations (months), first to last complete month", f"{T['n_months']} ({J['first']} to {J['last_complete']})"), 396 rows = [("Observations (months), first to last complete month", f"{T['n_months']} ({J['first']} to {J['last_complete']})"),
395 ("Placebo rank of the level $t$-statistic (all candidate months)", f"rank {T['placebo_chatgpt']['level_t_rank']} of {T['placebo_chatgpt']['n_placebo']}, placebo $p$ = {T['placebo_chatgpt']['level_t_placebo_p']:.3f}"), 397 ("Placebo rank of the level $t$-statistic (all candidate months)", f"rank {T['placebo_chatgpt']['level_t_rank']} of {T['placebo_chatgpt']['n_placebo']}, placebo $p$ = {T['placebo_chatgpt']['level_t_placebo_p']:.3f}"),
396 ("Placebo rank of the slope $t$-statistic (all candidate months)", f"rank {T['placebo_chatgpt']['slope_t_rank']} of {T['placebo_chatgpt']['n_placebo']}, placebo $p$ = {T['placebo_chatgpt']['slope_t_placebo_p']:.3f}"), 398 ("Placebo rank of the slope $t$-statistic (all candidate months)", f"rank {T['placebo_chatgpt']['slope_t_rank']} of {T['placebo_chatgpt']['n_placebo']}, placebo $p$ = {T['placebo_chatgpt']['slope_t_placebo_p']:.3f}"),
@@ -510,5 +512,51 @@ def tables_part09():
510 note="Within-cutoff AUC on the eligible topics of the cutoff; blank where the label window of the cutoff extends beyond the panel. Pooled values are in Tables~\\ref{tab:emerging_h12} and~\\ref{tab:emerging_h24}.") 512 note="Within-cutoff AUC on the eligible topics of the cutoff; blank where the label window of the cutoff extends beyond the panel. Pooled values are in Tables~\\ref{tab:emerging_h12} and~\\ref{tab:emerging_h24}.")
511 513
512 514
515+# ---------------- Part V (The Artificial Intelligence Transformation): tables appended by Agent E, 2026-11-03 ----------------
516+AI_CATS = ["llm", "foundation_models", "multimodal_ai", "generative_ai", "ai_agents", "reinforcement_learning", "ml_theory", "deep_learning_architectures", "ai_safety"]
517+AI_NAMES = {"llm": "Large language models", "foundation_models": "Foundation models", "multimodal_ai": "Multimodal AI", "generative_ai": "Generative AI", "ai_agents": "AI agents", "reinforcement_learning": "Reinforcement learning", "ml_theory": "ML theory", "deep_learning_architectures": "DL architectures", "ai_safety": "AI safety and alignment"}
518+
519+
520+def tables_part05():
521+ B = f"{R}/results/bibliometrics"; G = f"{R}/results/graph"; E = f"{R}/results/econometrics"; S = f"{R}/literature_synthesis"
522+ p = f"{B}/cagr_by_category.csv"
523+ if os.path.exists(p):
524+ d = pd.read_csv(p).drop_duplicates("category"); d = d[d.category.isin(AI_CATS)].set_index("category").loc[AI_CATS].reset_index()
525+ t = pd.DataFrame({"Category": d.category.map(AI_NAMES), "2018": d.n_2018, "2022": d.n_2022, "2023": d.n_2023, "2025": d.n_2025, "CAGR 2018--22": d.cagr_baseline, "CAGR 2023--25": d.cagr_post, "Share 2018 (%)": 100 * d.share_2018, "Share 2022 (%)": 100 * d.share_2022, "Share 2025 (%)": 100 * d.share_2025})
526+ write("tab_ai_growth", t, "Volume, compound annual growth and corpus share of the nine AI categories, baseline (2018--2022) and post (2023--2025) periods.", "tab:ai_growth", [p], fmt={"CAGR 2018--22": "{:.3f}", "CAGR 2023--25": "{:.3f}", "Share 2018 (%)": "{:.2f}", "Share 2022 (%)": "{:.2f}", "Share 2025 (%)": "{:.2f}"},
527+ note="Counts are multi-label (a work counts in every category it is assigned to), restricted to the eligible mapping corpus. CAGR as in Part IV, eq.~\\ref{eq:cagr}; the post-period CAGR uses 2023 and 2025 end points. Shares are over all eligible works of the year.")
528+ p = f"{G}/category_counts_yearly.csv"
529+ if os.path.exists(p):
530+ d = pd.read_csv(p); d = d[d.category_id.isin(AI_CATS) & (d.year <= 2026)].groupby(["category_id", "year"]).n_multilabel.sum().unstack("year").loc[AI_CATS]
531+ d.columns = [str(c) for c in d.columns]; d = d.reset_index(); d["category_id"] = d.category_id.map(AI_NAMES); d = d.rename(columns={"category_id": "Category"})
532+ write("tab_ai_counts", d, "Yearly multi-label counts of the nine AI categories (all period rows summed; 2026 incomplete).", "tab:ai_counts", [p], longtable=False, note="The source file splits each year into period rows (baseline / pre / post relative to 30 November 2022); they are summed here. 2026 covers publications dated up to the snapshot of 14 October 2026.")
533+ p = f"{E}/breaks.csv"
534+ if os.path.exists(p):
535+ d = pd.read_csv(p); d = d[d.series.isin(["cat_" + c for c in AI_CATS])].set_index("series").loc[["cat_" + c for c in AI_CATS]].reset_index()
536+ t = pd.DataFrame({"Category": d.series.str[4:].map(AI_NAMES), "Level 2022-12": d.chatgpt_level, "p": d.chatgpt_level_p, "Slope 2022-12": d.chatgpt_slope, "p ": d.chatgpt_slope_p, "Level 2023-04": d.gpt4_level, "p ": d.gpt4_level_p, "Level 2024-09": d.o1_level, "p ": d.o1_level_p, "Placebo p (level)": d.placebo_level_p, "Placebo p (slope)": d.placebo_slope_p, "Bai--Perron (BIC)": d.bp_breaks_bic.fillna("").astype(str).str.replace(";", "; "), "PELT": d.pelt.fillna("").astype(str).str.replace(";", "; ")})
537+ write("tab_ai_breaks", t, "Structural-change estimates for the nine AI categories: interrupted-time-series level and slope changes at 2022-12 (ChatGPT), level changes at the alternative event dates 2023-04 (GPT-4) and 2024-09 (o1), placebo p-values, and Bai--Perron and PELT break dates.", "tab:ai_breaks", [p], fmt={c: "{:.3f}" for c in ["Level 2022-12", "p", "Slope 2022-12", "p ", "Level 2023-04", "p ", "Level 2024-09", "p ", "Placebo p (level)", "Placebo p (slope)"]}, longtable=True,
538+ note="Monthly log counts, seasonally adjusted; HAC p-values (Part IV, Section~\\ref{sec:m-its}). Level changes are in log points of the monthly count; slope changes in log points per month. Placebo p-values are the share of placebo event dates (Section~\\ref{sec:m-placebo}) whose t-statistic exceeds the true one. Bai--Perron dates: breaks selected by BIC with trimming h=15 months; PELT: penalised exact linear-time change points. The categories are multi-label, so the series overlap.")
539+ p = f"{G}/category_flows_yearly.csv"
540+ if os.path.exists(p):
541+ d = pd.read_csv(p); cols = ["self", "llm", "nlp", "computer_vision", "hci", "robotics", "software_engineering", "ai_safety"]; rows = []
542+ for c in AI_CATS:
543+ for y in (2019, 2022, 2025):
544+ s = d[(d.citing_cat == c) & (d.citing_year == y)].set_index("cited_cat").row_share
545+ rows.append({"Category": AI_NAMES[c], "Year": str(y), "Self": 100 * s.get(c, 0), "LLM": 100 * s.get("llm", 0), "NLP": 100 * s.get("nlp", 0), "Vision": 100 * s.get("computer_vision", 0), "HCI": 100 * s.get("hci", 0), "Robotics": 100 * s.get("robotics", 0), "Softw. eng.": 100 * s.get("software_engineering", 0), "AI safety": 100 * s.get("ai_safety", 0)})
546+ t = pd.DataFrame(rows); write("tab_ai_flows", t, "Destination of the in-corpus references of the nine AI categories in 2019, 2022 and 2025 (per cent of in-corpus references of the citing category).", "tab:ai_flows", [p], fmt={c: "{:.1f}" for c in ["Self", "LLM", "NLP", "Vision", "HCI", "Robotics", "Softw. eng.", "AI safety"]}, longtable=True,
547+ note="Row shares of category-to-category citation flows (Part IV, eq.~\\ref{eq:flow}); references outside the eligible corpus are excluded. `Self' is the share going to the citing category itself (for LLM, the Self and LLM columns coincide). Categories are multi-label, so a reference can count under several destination categories.")
548+ p = f"{S}/agent_E/contributions.csv"
549+ if os.path.exists(p):
550+ d = pd.read_csv(p, dtype={"openalex_id": str})
551+ piv = pd.crosstab(d.primary_category.map(AI_NAMES), d.data_quality_flag).loc[[AI_NAMES[c] for c in AI_CATS]]
552+ piv["total"] = piv.sum(axis=1); piv = piv.reset_index().rename(columns={"primary_category": "Primary category"})
553+ write("tab_ai_landmark_flags", piv, "Specialist screening of the 234 AI/ML landmark candidates: records by primary category and data-quality flag.", "tab:ai_landmark_flags", [p], note="Flags assigned by Agent E from title, venue, identifiers, citation counts and (where available) the abstract: identifier\\_mismatch = title/abstract do not correspond to the arXiv identifier or citation count attached to the record; abstract\\_mismatch = stored abstract belongs to another work; year\\_implausible = OpenAlex year contradicts the first publication; normalisation\\_artefact = very small c36 with very large NCS36; off\\_scope = not an AI contribution; no\\_abstract = metadata only. `ok' records were judged from the abstract.")
554+ ct = pd.crosstab(d.primary_category.map(AI_NAMES), d.contribution_type).loc[[AI_NAMES[c] for c in AI_CATS]]; ct["total"] = ct.sum(axis=1); ct = ct.reset_index().rename(columns={"primary_category": "Primary category"})
555+ write("tab_ai_contribution_types", ct, "Contribution type of the 234 AI/ML landmark candidates by primary category (specialist judgement from abstract or metadata).", "tab:ai_contribution_types", [p], note="Types follow CLAUDE.md~\\S1 item 6: theory, method, empirical, benchmark, engineering, survey. Judged from the abstract for 190 records and from title/venue/type for 44 records without abstract (column judged\\_from of the source file).")
556+ f = d[["openalex_id", "year", "type", "primary_category", "c36", "ncs36", "contribution_type", "data_quality_flag"]].copy(); f["primary_category"] = f.primary_category.map(AI_NAMES); f["c36"] = f.c36.astype(int); f["year"] = f.year.astype(str)
557+ f.columns = ["OpenAlex id", "Year", "Type", "Category", "c36", "NCS36", "Contribution", "Flag"]
558+ write("tab_ai_landmarks", f, "The 234 AI/ML landmark candidates with the specialist's contribution type and data-quality flag (titles in literature_synthesis/agent_E/contributions.csv).", "tab:ai_landmarks", [p], fmt={"NCS36": "{:.1f}"}, longtable=True, note="Selection score and windows as in Part IV (Section~\\ref{sec:m-score}); c36 and NCS36 as stored in the landmark file.")
559+
560+
513 if __name__ == "__main__": 561 if __name__ == "__main__":
514 main() 562 main()
515 563
Menu

Theme
↑↓ navigate ↵ open esc close / or ⌘K anywhere