{"id":"W4286449976","doi":"10.18280/ria.360307","title":"A New Supervised Term Weight Measure Based Approach for Text Classification","year":2022,"lang":"en","type":"article","venue":"Revue d intelligence artificielle","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Term (time); Computer science; Measure (data warehouse); Task (project management); Identification (biology); Support vector machine; tf–idf; Categorization; Text categorization; Artificial intelligence; Information retrieval; Document classification; Natural language processing; Pattern recognition (psychology); Data mining","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001778337,0.001030805,0.001524008,0.007280501,0.0008315147,0.001439908,0.001565061,0.001246374,0.001611121],"category_scores_gemma":[0.005680545,0.0001974176,0.001250771,0.00576262,0.0005989733,0.002911154,0.0007745085,0.001262025,0.001334585],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001199763,"about_ca_system_score_gemma":0.001354024,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003067092,"about_ca_topic_score_gemma":0.003315391,"domain_scores_codex":[0.9971676,0.0003341863,0.0003560357,0.0005366195,0.001456636,0.0001488731],"domain_scores_gemma":[0.9977263,0.000651797,0.0002721033,0.0001588219,0.001107045,0.00008391318],"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":[0.0003727842,0.0003730632,0.004458298,0.0003506321,0.0001802966,0.0001293567,0.0001671574,0.01230061,0.02495409,0.004617457,0.008857758,0.9432386],"study_design_scores_gemma":[0.00008512935,0.0007265888,0.01014763,0.0001109725,0.0002079924,0.0009673075,0.0002174211,0.9256335,0.02662829,0.01361872,0.02152233,0.0001342275],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04034065,0.003113014,0.9488099,0.0003575428,0.0004165545,0.000380559,0.001048399,0.002167991,0.003365414],"genre_scores_gemma":[0.3790884,0.001534368,0.6037347,0.0003393202,0.0007784974,0.0009518658,0.005074652,0.000271759,0.008226325],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007280501,"threshold_uncertainty_score":0.009404838,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07099948823014113,"score_gpt":0.2729647282679941,"score_spread":0.201965240037853,"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."}}