{"id":"W2010920802","doi":"10.4028/www.scientific.net/amm.599-601.1824","title":"Comparison and Improvements of Feature Extraction Methods for Text Categorization","year":2014,"lang":"en","type":"article","venue":"Applied Mechanics and Materials","topic":"Advanced Computational Techniques and Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Advanced Micro Devices (Canada)","funders":"","keywords":"Categorization; Text categorization; Computer science; Feature extraction; Feature (linguistics); Point (geometry); Artificial intelligence; Pattern recognition (psychology); Mutual information; Key (lock); Information extraction; Data mining; 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":[0.004926184,0.001450566,0.001384278,0.006681411,0.0005845808,0.001171951,0.001522224,0.001028845,0.002211887],"category_scores_gemma":[0.01386523,0.0003158796,0.001442303,0.005210458,0.0004175344,0.003791964,0.0005967993,0.001062519,0.001555703],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007237804,"about_ca_system_score_gemma":0.0006539514,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001751029,"about_ca_topic_score_gemma":0.001463284,"domain_scores_codex":[0.99428,0.001253723,0.0006194611,0.0008540846,0.00277799,0.0002147425],"domain_scores_gemma":[0.9890659,0.005550358,0.0005492934,0.0009664245,0.003727539,0.0001403677],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.000297122,0.0001345953,0.002667792,0.000584675,0.0001921863,0.00004925747,0.0001052526,0.002665241,0.01075388,0.0009104375,0.003878362,0.9777613],"study_design_scores_gemma":[0.0004168974,0.004594601,0.128931,0.0007502737,0.00199739,0.003156643,0.001042484,0.5401348,0.182019,0.01796363,0.1181915,0.0008018163],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.08663135,0.04687004,0.8500108,0.0008943134,0.001231355,0.0006842437,0.001484788,0.006374508,0.005818682],"genre_scores_gemma":[0.2849045,0.01082026,0.6946115,0.0002461451,0.0009157803,0.0006640179,0.003296261,0.0003686193,0.004172908],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006681411,"threshold_uncertainty_score":0.02605247,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02029186270272505,"score_gpt":0.3669829633220958,"score_spread":0.3466911006193708,"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."}}