{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00003517041,0.00004241969,0.00003783708,0.00001747923,0.00006973233,0.0001386337,0.0003805086,0.000007522528,0.0001033359],"category_scores_gemma":[9.858884e-7,0.00002974705,0.000007516062,0.0001230795,0.00001242879,0.0002465005,0.000127591,0.00002904317,0.0001882323],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000006875153,"about_ca_system_score_gemma":0.000001913609,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002511119,"about_ca_topic_score_gemma":0.00001349368,"domain_scores_codex":[0.9996269,0.000004223818,0.000053823,0.0001356882,0.00008409831,0.00009529838],"domain_scores_gemma":[0.9995708,0.00001427551,0.00001423889,0.0003529147,0.00001142779,0.0000362935],"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":[3.963536e-7,0.00007857569,0.0002510344,0.000004703493,0.00001370037,0.000007714984,0.0006664769,0.0001201183,0.00005295154,0.119035,0.03357806,0.8461913],"study_design_scores_gemma":[0.0002384458,0.00008200607,0.0007999474,0.00001194717,0.000002459991,0.00003540661,0.00004692666,0.8317521,0.0003950968,0.0005360868,0.1659216,0.0001779288],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0007027647,0.00001998423,0.9627187,0.002543746,0.00002376396,0.00004199128,5.814743e-7,0.0001446093,0.03380391],"genre_scores_gemma":[0.202933,0.00001000977,0.7916515,0.0005582559,0.00002541155,0.00002362711,0.000001063166,0.000003663221,0.004793514],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8460134,"threshold_uncertainty_score":0.2419408,"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."}}