{"id":"W4254656533","doi":"10.1007/978-1-4614-6170-8_100587","title":"Categorical Data Similarity","year":2014,"lang":"en","type":"book-chapter","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Calgary","funders":"","keywords":"Categorical variable; Similarity (geometry); Artificial intelligence; Computer science; Natural language processing; Mathematics; Statistics","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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0002347603,0.0001954431,0.0002199682,0.0000524891,0.00009550829,0.0002045126,0.004473574,0.0001791484,0.0001552273],"category_scores_gemma":[0.00001654932,0.0001692341,0.00004044922,0.00003168096,0.00004758754,0.0001977595,0.002676438,0.0002825978,0.001798745],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002004506,"about_ca_system_score_gemma":0.00008032912,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002682922,"about_ca_topic_score_gemma":0.00001538361,"domain_scores_codex":[0.9985189,0.000006054358,0.000217434,0.0008318574,0.0002569365,0.0001688152],"domain_scores_gemma":[0.995454,0.00008115646,0.00009261347,0.004203923,0.00005101708,0.0001173224],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[7.120046e-8,0.000004041035,1.61429e-7,0.000002685527,0.000007201318,0.000002668384,0.000002914352,1.611204e-7,2.141965e-7,0.7130093,0.182704,0.1042666],"study_design_scores_gemma":[0.00003619975,0.00001131711,0.000003803406,0.000005402573,0.00001019258,0.00001205364,2.045178e-7,0.06402738,0.000001355983,0.05868812,0.8770016,0.0002023774],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"other","genre_scores_codex":[1.046241e-8,0.00003315798,0.5466482,0.0009266456,0.0001084198,0.00005156608,0.00007978743,0.0001452792,0.4520069],"genre_scores_gemma":[0.00003742586,0.00005382474,0.2120298,0.0007577987,0.0003107815,0.000004833556,0.001160038,0.00002007837,0.7856255],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.6942976,"threshold_uncertainty_score":0.9989785,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08237857441381383,"score_gpt":0.2816371582400747,"score_spread":0.1992585838262609,"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."}}