{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007348122,0.001300108,0.001926369,0.006747143,0.004493394,0.01185228,0.005488054,0.002007262,0.02574909],"category_scores_gemma":[0.02896587,0.001079906,0.001858622,0.01112632,0.003111491,0.0216034,0.01230947,0.002829982,0.01671588],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002694823,"about_ca_system_score_gemma":0.005066921,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006166916,"about_ca_topic_score_gemma":0.00598308,"domain_scores_codex":[0.9873761,0.002189403,0.001575034,0.002098649,0.005729221,0.001031629],"domain_scores_gemma":[0.9675607,0.005505349,0.001411055,0.01812858,0.006579311,0.0008149689],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004227409,0.0001704068,0.002198852,0.0006948376,0.00009587537,0.0004092358,0.002140884,0.003979548,0.006805926,0.4884667,0.133655,0.3609599],"study_design_scores_gemma":[0.00002962936,0.00005661768,0.0003948597,0.0003123895,0.00005841224,0.0005547975,0.000693325,0.0151334,0.01091038,0.2487431,0.7229906,0.0001225105],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004146572,0.00145741,0.9173707,0.001413068,0.001032582,0.0004774136,0.00625408,0.01329827,0.05454989],"genre_scores_gemma":[0.09355291,0.002994275,0.8080494,0.002776547,0.001124723,0.001035734,0.02923199,0.008889377,0.05234508],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02574909,"threshold_uncertainty_score":0.08613926,"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."}}