{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001496095,0.0008696031,0.0010782,0.007984398,0.001449708,0.002246083,0.001855345,0.001292457,0.003598521],"category_scores_gemma":[0.01201676,0.0002906509,0.001148359,0.008079923,0.0007131147,0.002033833,0.001055905,0.0008664334,0.003460889],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008520809,"about_ca_system_score_gemma":0.001294191,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002069553,"about_ca_topic_score_gemma":0.00223606,"domain_scores_codex":[0.9965333,0.0005663297,0.0003863746,0.0007451235,0.001570301,0.0001986494],"domain_scores_gemma":[0.9951512,0.001627555,0.0004928025,0.0007643718,0.001852032,0.0001119756],"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":[0.0005075346,0.0002312406,0.00570504,0.001271737,0.0002414935,0.0008774958,0.00066431,0.05211028,0.0242716,0.05551638,0.02625808,0.8323449],"study_design_scores_gemma":[0.000114471,0.0002790494,0.006997007,0.0002734619,0.0002244559,0.003565968,0.0006298505,0.6956286,0.04418018,0.1643912,0.08349011,0.0002255621],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02027089,0.001062256,0.9664925,0.0002577996,0.000176737,0.0006484244,0.00366284,0.002608807,0.004819858],"genre_scores_gemma":[0.1316477,0.0007546039,0.8527367,0.0001297142,0.0001717081,0.0005633195,0.009833001,0.0001950017,0.003968327],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007984398,"threshold_uncertainty_score":0.01203823,"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."}}