{"id":"W4410546712","doi":"10.18280/isi.300408","title":"An Accurate Model for Text Document Classification Using Machine Learning Techniques","year":2025,"lang":"en","type":"article","venue":"Ingénierie des systèmes d information","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Artificial intelligence; Machine learning; Document classification; Natural language processing; Information retrieval","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":[0.001744902,0.0009601241,0.001230803,0.002971665,0.0005324307,0.0018913,0.002015074,0.001737854,0.001489404],"category_scores_gemma":[0.004797272,0.0003414319,0.001160946,0.002659532,0.0004493731,0.00294288,0.0006442048,0.001849263,0.00293441],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001383693,"about_ca_system_score_gemma":0.001149423,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004952286,"about_ca_topic_score_gemma":0.004209292,"domain_scores_codex":[0.9986346,0.0002671745,0.0001513141,0.0003809233,0.0004665294,0.00009939692],"domain_scores_gemma":[0.9981324,0.0008495244,0.0001731297,0.0002006523,0.0006052665,0.00003899397],"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":[0.0003078007,0.0004267215,0.004902762,0.000475578,0.0001472663,0.00032779,0.0002320914,0.3614134,0.0109188,0.01771246,0.01393561,0.5891997],"study_design_scores_gemma":[0.000006214772,0.00004100874,0.0004281575,0.00002332433,0.00001785292,0.00008018065,0.00001398243,0.9912308,0.00125816,0.004252664,0.002636054,0.00001160535],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01963708,0.001544752,0.9721054,0.0007968124,0.0002912209,0.0002616352,0.001026853,0.002793289,0.001542976],"genre_scores_gemma":[0.3509316,0.002478368,0.6311946,0.000451948,0.0005698086,0.001045741,0.003860049,0.0001964269,0.009271519],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004952286,"threshold_uncertainty_score":0.01003945,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03506586237580583,"score_gpt":0.304176570138292,"score_spread":0.2691107077624861,"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."}}