{"id":"W7130592618","doi":"10.1109/fllm67465.2025.11391003","title":"Automated Research Article Classification and Recommendation Using NLP and Machine Learning","year":2025,"lang":"","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Algoma University; York University","funders":"","keywords":"Cosine similarity; Support vector machine; Random forest; Information overload; Complement (music); Statistical classification; Feature extraction; Feature (linguistics)","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":[0.003036176,0.001219131,0.002016891,0.02003377,0.001365619,0.002826228,0.001777774,0.001434196,0.003306221],"category_scores_gemma":[0.01580097,0.0005002627,0.001350575,0.01292328,0.0004196547,0.003453579,0.00107452,0.001123213,0.008706959],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001328405,"about_ca_system_score_gemma":0.003416173,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01476259,"about_ca_topic_score_gemma":0.02647655,"domain_scores_codex":[0.9956995,0.0008206659,0.000528673,0.001000818,0.001692817,0.0002575146],"domain_scores_gemma":[0.9863751,0.004738376,0.001959308,0.002080869,0.004375439,0.0004709305],"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":[0.0002832674,0.0005214092,0.01927458,0.0007529747,0.0002008337,0.0003052214,0.0001723005,0.005831788,0.01818756,0.002240303,0.05802648,0.8942032],"study_design_scores_gemma":[0.0003115506,0.0004661194,0.03055286,0.0002929532,0.0003409346,0.001276093,0.0007739754,0.7457172,0.0757379,0.02865128,0.1156601,0.0002189704],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09867315,0.01042889,0.7166359,0.00576131,0.0007558769,0.002540623,0.05048786,0.0953982,0.01931816],"genre_scores_gemma":[0.1794425,0.002092261,0.7555246,0.0005348682,0.0006275048,0.0006149189,0.0530091,0.0004780935,0.007676232],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02003377,"threshold_uncertainty_score":0.02935332,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1414599886026211,"score_gpt":0.4417872786589687,"score_spread":0.3003272900563476,"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."}}