{"id":"W2117423052","doi":"10.1109/icdm.2002.1183881","title":"Text document categorization by term association","year":2003,"lang":"en","type":"article","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":224,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Classifier (UML); Categorization; Artificial intelligence; Text categorization; Association rule learning; Machine learning; Natural language processing; Data mining; Information retrieval; Pattern recognition (psychology)","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":[0.001965918,0.001288859,0.002104377,0.01887806,0.001359844,0.003091806,0.001630575,0.001345324,0.003847195],"category_scores_gemma":[0.008328383,0.0003848431,0.001952354,0.01496377,0.0009172916,0.003989653,0.001470869,0.001343699,0.00651803],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008152172,"about_ca_system_score_gemma":0.001645116,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002096726,"about_ca_topic_score_gemma":0.002086104,"domain_scores_codex":[0.9969261,0.0004566119,0.000495104,0.0008055436,0.00115198,0.0001646471],"domain_scores_gemma":[0.996585,0.001324837,0.0004781901,0.0005127912,0.000969644,0.0001295363],"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":[0.0001850968,0.0002342459,0.002929035,0.0008393743,0.0001679322,0.0001979212,0.0002277342,0.003810916,0.01687346,0.008394107,0.01969717,0.946443],"study_design_scores_gemma":[0.0002499652,0.000923178,0.01644896,0.0007967952,0.000677807,0.005677599,0.001193084,0.4947587,0.07439812,0.1688602,0.2355898,0.0004257906],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02531536,0.004801477,0.947479,0.0006406003,0.0005637011,0.002040956,0.005908788,0.007241863,0.006008335],"genre_scores_gemma":[0.04351626,0.001363272,0.9436159,0.0001143968,0.0002554178,0.0009065357,0.006849216,0.0002173185,0.003161681],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01887806,"threshold_uncertainty_score":0.01287013,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005463032048834978,"score_gpt":0.2254333012677956,"score_spread":0.2199702692189607,"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."}}