{"id":"W4210572593","doi":"10.1145/775107.775110","title":"Frequent term-based text clustering","year":2002,"lang":"en","type":"article","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Cluster analysis; Computer science; Term (time); Document clustering; Data mining; Information retrieval; Clustering high-dimensional data; Hierarchical clustering; Artificial intelligence","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":[0.00004958308,0.00005588442,0.00004773877,0.0000337829,0.00007810449,0.0001372552,0.0005305794,0.0000164953,0.0002270836],"category_scores_gemma":[0.000004521081,0.00004776987,0.00002186921,0.0001521491,0.00001379625,0.0001750526,0.0001065847,0.00003998465,0.0005451016],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001342646,"about_ca_system_score_gemma":0.000005425134,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002055816,"about_ca_topic_score_gemma":0.00000439483,"domain_scores_codex":[0.9994772,0.000006579661,0.00009079411,0.0001993583,0.00009361067,0.0001324696],"domain_scores_gemma":[0.9994025,0.00002239854,0.00002066357,0.0004849962,0.00001442274,0.00005506805],"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":[1.604653e-7,0.0001129375,0.0001843858,0.000005974563,0.000004866757,0.000006786729,0.0001965202,0.0001580146,0.0008350352,0.01671469,0.02084212,0.9609385],"study_design_scores_gemma":[0.0001026768,0.00001504513,0.0005585409,0.000005393425,9.145978e-7,0.000004058446,0.000004067871,0.9713961,0.0006347137,0.0001051622,0.02708853,0.00008480862],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0007565757,0.00002554857,0.9746438,0.002045792,0.00006974447,0.00004985269,0.000002122579,0.0002088418,0.02219778],"genre_scores_gemma":[0.3337743,0.000004661803,0.6623369,0.001275097,0.00005002885,0.000031074,0.000003502056,0.000005787205,0.002518628],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9712381,"threshold_uncertainty_score":0.700636,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03268005172658556,"score_gpt":0.2419656934351582,"score_spread":0.2092856417085726,"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."}}