{"id":"W2145753452","doi":"10.1109/ccece.2007.203","title":"Document Classification with ACM Subject Hierarchy","year":2007,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Computer science; Categorization; Information retrieval; Hierarchy; Classifier (UML); Document classification; Text categorization; Digital library; Classification scheme; Library classification; Subject (documents); Focus (optics); Context (archaeology); Artificial intelligence; World Wide Web; Natural language processing","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":[0.005641483,0.00199992,0.002425045,0.02550891,0.002514459,0.007795663,0.002457761,0.002044053,0.04946832],"category_scores_gemma":[0.02269102,0.0007242068,0.003076999,0.02986802,0.0009241817,0.005223043,0.003417768,0.002590082,0.04849719],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001833902,"about_ca_system_score_gemma":0.006290942,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0114861,"about_ca_topic_score_gemma":0.01085243,"domain_scores_codex":[0.9904782,0.00168028,0.002265174,0.001804452,0.003114268,0.0006576383],"domain_scores_gemma":[0.9894302,0.002724122,0.001023252,0.002810913,0.003594371,0.0004171287],"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":[0.0002065475,0.0001890422,0.002272911,0.001297893,0.0001991811,0.0002146794,0.0002329961,0.003398157,0.00309788,0.01190323,0.1978571,0.7791303],"study_design_scores_gemma":[0.0003224899,0.0005104474,0.00888156,0.0007924095,0.0003068435,0.001694894,0.0004485116,0.1468122,0.009973961,0.07713405,0.7529047,0.0002180321],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.007820886,0.01089998,0.7417315,0.002656818,0.002774178,0.01086073,0.1047111,0.0778316,0.0407132],"genre_scores_gemma":[0.0502894,0.003991799,0.8173311,0.0005699198,0.001462936,0.006124581,0.08523712,0.001759043,0.03323403],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.04946832,"threshold_uncertainty_score":0.165488,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02711306376552129,"score_gpt":0.2802209110725788,"score_spread":0.2531078473070575,"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."}}