{"id":"W2154923335","doi":"10.1109/compsac.2005.128","title":"Recovering \"Lack of Words\" in Text Categorization for Item Banks","year":2006,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Categorization; Computer science; Phrase; Text categorization; Feature selection; Artificial intelligence; Classifier (UML); Natural language processing; Support vector machine; Selection (genetic algorithm); Sample (material); Machine learning","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":[0.002889975,0.001105646,0.001404115,0.002818541,0.0007792525,0.001071666,0.001015828,0.001065268,0.001347808],"category_scores_gemma":[0.01450652,0.0003328603,0.000950503,0.003791409,0.0007908631,0.003407938,0.001426781,0.001391398,0.001483848],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003967909,"about_ca_system_score_gemma":0.0008257,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001028184,"about_ca_topic_score_gemma":0.000939982,"domain_scores_codex":[0.9972174,0.0008021333,0.0004459157,0.0005065905,0.0008292549,0.0001986455],"domain_scores_gemma":[0.9892097,0.006058687,0.001236524,0.001502158,0.001782097,0.0002109054],"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":[0.000770987,0.0003711038,0.02142205,0.000619493,0.0001542249,0.0005978536,0.0007741838,0.009474849,0.04705835,0.002204725,0.007502059,0.9090501],"study_design_scores_gemma":[0.000197903,0.002375977,0.06630684,0.000202556,0.0004547779,0.003403597,0.002459911,0.7359308,0.1304692,0.04078119,0.01719828,0.0002190198],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2231063,0.0005553079,0.7706384,0.0003972609,0.000115035,0.0004052627,0.001093838,0.002303565,0.001385106],"genre_scores_gemma":[0.6960121,0.0003151581,0.2958396,0.0002030982,0.0002078467,0.0006571011,0.004631377,0.0002036805,0.001929905],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002889975,"threshold_uncertainty_score":0.01528382,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03400157365506067,"score_gpt":0.277484185770742,"score_spread":0.2434826121156813,"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."}}