{"id":"W4413364391","doi":"10.1016/j.procs.2025.07.178","title":"Retrieve-Classify-Read: Passage Filtering via Subject Classification for University Question Answering","year":2025,"lang":"en","type":"article","venue":"Procedia Computer Science","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Acadia University","funders":"","keywords":"Computer science; Question answering; Subject (documents); Information retrieval; Artificial intelligence; Natural language processing; World Wide Web","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":[],"consensus_categories":[],"category_scores_codex":[0.001077162,0.0001673711,0.0001658804,0.0004095041,0.0005202951,0.0003626528,0.001887564,0.00007223482,0.000001306169],"category_scores_gemma":[0.0001123211,0.0001828513,0.00005836455,0.001745518,0.0001537242,0.001586924,0.0006078727,0.0001471805,0.00000627995],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003150526,"about_ca_system_score_gemma":0.0005486072,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002137236,"about_ca_topic_score_gemma":0.00000789273,"domain_scores_codex":[0.9979553,0.00003760822,0.0002438748,0.0009517256,0.0003553505,0.0004561857],"domain_scores_gemma":[0.9986188,0.0001455128,0.0001213827,0.0006592206,0.0003399292,0.0001151003],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002651021,0.00007916389,0.003023515,0.0001985892,0.00001444572,0.000008189743,0.0009256514,0.001981045,0.03003012,0.3303431,0.0003132292,0.6330565],"study_design_scores_gemma":[0.0002654884,0.00004817498,0.007637253,0.0000848134,0.00000535804,0.000006257541,0.00001217871,0.9799473,0.004630088,0.004541536,0.002622297,0.0001992971],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01080075,0.00002600763,0.9849945,0.001180189,0.001287044,0.0003640458,0.000001455194,0.0003863206,0.000959719],"genre_scores_gemma":[0.6204981,0.000006016237,0.3790532,0.0001549104,0.00008878338,0.000008341882,0.000001333127,0.000004567965,0.0001847342],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9779662,"threshold_uncertainty_score":0.7456459,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02297700179631848,"score_gpt":0.2571357960003268,"score_spread":0.2341587942040083,"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."}}