{"id":"W2107131706","doi":"10.1109/icdm.2006.64","title":"Enhancing Text Clustering Using Concept-based Mining Model","year":2006,"lang":"en","type":"article","venue":"Proceedings","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":61,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Cluster analysis; Phrase; Sentence; Document clustering; Term (time); Semantics (computer science); Similarity (geometry); Natural language processing; Matching (statistics); Artificial intelligence; Similarity measure; Measure (data warehouse); Word (group theory); Information retrieval; Data mining; Mathematics; Image (mathematics)","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":[0.002543673,0.0009222404,0.00152239,0.003132153,0.000723102,0.001466218,0.001887662,0.001246,0.001124006],"category_scores_gemma":[0.009291006,0.0003717512,0.001165754,0.003664048,0.0005518639,0.00331083,0.00109646,0.000997918,0.0008760668],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008505969,"about_ca_system_score_gemma":0.001251326,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00206702,"about_ca_topic_score_gemma":0.002152088,"domain_scores_codex":[0.9981452,0.0004657971,0.0001217934,0.0003266412,0.0008691802,0.00007133373],"domain_scores_gemma":[0.9959488,0.002042172,0.0002481069,0.0003281452,0.001367261,0.0000654908],"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":[0.0003824224,0.0008514712,0.004025596,0.0007997977,0.0003822511,0.0003963237,0.0008550308,0.1951983,0.03362295,0.02601971,0.007341005,0.7301251],"study_design_scores_gemma":[0.00003807523,0.0000990854,0.0005861864,0.00002483846,0.00007191855,0.0002761485,0.0001091435,0.971871,0.01034535,0.01296462,0.00357489,0.00003877798],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01794225,0.0002548921,0.9792842,0.0001721745,0.00004398574,0.0001875405,0.00009928611,0.0008893226,0.001126389],"genre_scores_gemma":[0.1177167,0.0003174047,0.8800128,0.0001165373,0.00006641315,0.0003179471,0.0004112952,0.00007115544,0.0009696566],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003132153,"threshold_uncertainty_score":0.01345241,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02220185227204753,"score_gpt":0.2452552487338726,"score_spread":0.223053396461825,"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."}}