{"id":"W4245642361","doi":"10.1109/ideas.2007.4318085","title":"An Approach for Text Categorization in Digital Library","year":2007,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Categorization; Computer science; Hierarchy; Text categorization; Information retrieval; Digital library; Document classification; Library classification; Artificial intelligence; 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":[0.001251697,0.0006428411,0.0005883609,0.004402564,0.002038433,0.001943525,0.001688582,0.001345786,0.003175165],"category_scores_gemma":[0.003670692,0.0003052692,0.001261538,0.004323729,0.0009845382,0.002212237,0.00157796,0.001426509,0.002285984],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001617962,"about_ca_system_score_gemma":0.001844934,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008209468,"about_ca_topic_score_gemma":0.0115194,"domain_scores_codex":[0.9975745,0.0005437253,0.0001799826,0.0004999937,0.001060363,0.0001414525],"domain_scores_gemma":[0.9986143,0.0004334453,0.0001072199,0.0002339955,0.0005249257,0.00008600295],"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":[0.0001078922,0.0002835893,0.002001315,0.000318531,0.00006562562,0.0002112584,0.0009430245,0.007219503,0.01991834,0.02692743,0.01264937,0.9293541],"study_design_scores_gemma":[0.0001202927,0.0004923939,0.009036628,0.0002365541,0.0002871607,0.002045243,0.001741361,0.5868846,0.04618403,0.141788,0.2109991,0.0001845229],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009849273,0.0007096164,0.9818902,0.0006592786,0.0001497739,0.0006993561,0.0003421508,0.001636288,0.004064078],"genre_scores_gemma":[0.04568258,0.0003267803,0.9452246,0.0002717446,0.00009785232,0.0006047299,0.000651303,0.00006873927,0.007071557],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008209468,"threshold_uncertainty_score":0.01632339,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01624459706194404,"score_gpt":0.2479245301932853,"score_spread":0.2316799331313413,"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."}}