{"id":"W4404401204","doi":"10.54195/irrj.19910","title":"Annotative Indexing","year":2025,"lang":"en","type":"preprint","venue":"Information Retrieval Research","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Tsinghua University; University of Glasgow","keywords":"Search engine indexing; Computer science; Information retrieval","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["research_integrity"],"consensus_categories":[],"category_scores_codex":[0.003554692,0.000222236,0.0003216785,0.002311126,0.0003065796,0.001025039,0.002947693,0.0003974822,0.00002240579],"category_scores_gemma":[0.002664719,0.000220789,0.0001466131,0.00292236,0.0001457987,0.002737742,0.006042666,0.002444451,0.000213985],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005959448,"about_ca_system_score_gemma":0.001158859,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005718113,"about_ca_topic_score_gemma":0.000002691161,"domain_scores_codex":[0.9959745,0.0003386268,0.0006988129,0.000395354,0.002059265,0.0005334219],"domain_scores_gemma":[0.9947589,0.0006633488,0.0002934341,0.001377565,0.002785025,0.0001217209],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001161373,0.00007724597,0.0001913746,0.0007700101,0.0002131192,0.00001699317,0.01019599,0.003805096,0.0001140007,0.5616689,0.0204239,0.4024072],"study_design_scores_gemma":[0.0004839082,0.0001671766,0.0007221558,0.0008085148,0.00001564384,0.000003983932,0.0004310615,0.351374,0.03142219,0.4366107,0.1770799,0.0008808248],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0005260865,0.0001036954,0.9291891,0.001165175,0.000219898,0.0006821418,0.00002859341,0.0005410636,0.06754427],"genre_scores_gemma":[0.4999946,0.001065653,0.4861001,0.001586719,0.0003118099,0.0004215717,0.0004802423,0.00003203596,0.01000727],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.4994685,"threshold_uncertainty_score":0.9998569,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08469652661856103,"score_gpt":0.4552504328181435,"score_spread":0.3705539061995825,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}