{"id":"W2103986443","doi":"10.1145/564376.564412","title":"Document clustering with committees","year":2002,"lang":"en","type":"article","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":174,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Cluster analysis; Information retrieval; Artificial intelligence","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01101068,0.001411087,0.001850168,0.00524215,0.001829945,0.00392261,0.002203913,0.001941019,0.005017842],"category_scores_gemma":[0.04172497,0.0008146841,0.002098172,0.007240537,0.001158686,0.00413127,0.002873637,0.001880964,0.004063326],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001181089,"about_ca_system_score_gemma":0.001550262,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002179129,"about_ca_topic_score_gemma":0.002101592,"domain_scores_codex":[0.9838816,0.008081938,0.001405824,0.002474466,0.003717137,0.0004390365],"domain_scores_gemma":[0.9791988,0.008494368,0.00129603,0.005504679,0.005045365,0.0004606616],"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":[0.0006019738,0.0001560754,0.00273231,0.0006054219,0.0004244772,0.0001321603,0.0005595175,0.1352077,0.006054404,0.06102528,0.02604021,0.7664605],"study_design_scores_gemma":[0.0001395248,0.0003340204,0.001819139,0.0001172049,0.0001968932,0.0005211554,0.0002258051,0.8035221,0.0189684,0.1157939,0.05825046,0.0001115238],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005525586,0.0005929355,0.9872634,0.0001423815,0.0001015973,0.0003529771,0.0004202763,0.002164887,0.003435979],"genre_scores_gemma":[0.07907792,0.000468123,0.9120487,0.0001350357,0.00023216,0.0006482719,0.0023068,0.000647557,0.00443537],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01101068,"threshold_uncertainty_score":0.0582307,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01839922667157008,"score_gpt":0.2197457956381992,"score_spread":0.2013465689666291,"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."}}