{"id":"W2133411495","doi":"10.1109/ideas.2007.11","title":"An Approach for Text Categorization in Digital Library","year":2007,"lang":"en","type":"article","venue":"International Database Engineering and Applications Symposium","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Categorization; Computer science; Hierarchy; Digital library; Text categorization; Information retrieval; Library classification; Document classification; Artificial intelligence; Data mining; Machine learning; World Wide Web","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":[0.001185952,0.0006201764,0.0005632873,0.004509604,0.001981307,0.001994467,0.001644591,0.001325484,0.003182222],"category_scores_gemma":[0.003379587,0.0003056821,0.001250791,0.00441367,0.0009178765,0.002078474,0.0015781,0.001389013,0.002352988],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001607101,"about_ca_system_score_gemma":0.001862946,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008264404,"about_ca_topic_score_gemma":0.01189362,"domain_scores_codex":[0.9976318,0.000494449,0.0001747872,0.0004918276,0.001068846,0.000138347],"domain_scores_gemma":[0.9987214,0.000361172,0.00009889929,0.0002232495,0.0005125651,0.00008260831],"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":[0.00009987411,0.0002758417,0.002092118,0.0003174268,0.00006622943,0.0002150476,0.000910478,0.006869336,0.01972438,0.02791458,0.01414051,0.9273742],"study_design_scores_gemma":[0.0001074092,0.0004630124,0.009511322,0.0002488057,0.0002841413,0.002237946,0.001771392,0.5634345,0.04426388,0.1346319,0.242862,0.0001836311],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009111605,0.0007272105,0.9822813,0.0006552429,0.0001578346,0.0006901019,0.000361927,0.001678276,0.0043364],"genre_scores_gemma":[0.04321841,0.0003384578,0.9472541,0.0002756894,0.00009602368,0.0005731301,0.0006705797,0.00007013193,0.007503653],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008264404,"threshold_uncertainty_score":0.01643258,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.007484145434688436,"score_gpt":0.2315351532848341,"score_spread":0.2240510078501457,"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."}}