{"id":"W2143085379","doi":"","title":"Efficiently answering top-k typicality queries on large databases","year":2007,"lang":"en","type":"article","venue":"","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":34,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Computer science; Discriminative model; Set (abstract data type); Tree (set theory); Measure (data warehouse); Quality (philosophy); Information retrieval; Data mining; Artificial intelligence; Mathematics","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":[0.003578479,0.001099496,0.00293924,0.002170656,0.001670242,0.003692213,0.002813852,0.001988517,0.002101168],"category_scores_gemma":[0.02414353,0.00083337,0.001681373,0.004938206,0.001276841,0.009017266,0.002567123,0.001770098,0.0007251425],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001258491,"about_ca_system_score_gemma":0.001387727,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002808772,"about_ca_topic_score_gemma":0.003399661,"domain_scores_codex":[0.9944335,0.001616737,0.0006598864,0.001640765,0.001255939,0.0003931042],"domain_scores_gemma":[0.9800819,0.01363093,0.0014847,0.003174845,0.001135686,0.0004919309],"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":[0.001788001,0.0008173622,0.03278739,0.001423654,0.0006765258,0.0008499789,0.002416708,0.3626529,0.02096277,0.0730093,0.02497681,0.4776386],"study_design_scores_gemma":[0.00007830827,0.0001311651,0.001995524,0.00002358647,0.00006323461,0.0006010546,0.0006501779,0.8637142,0.004050347,0.1261424,0.002516089,0.00003397521],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2937954,0.001266388,0.6946444,0.002284178,0.00006732935,0.0003125689,0.002206456,0.002966848,0.002456374],"genre_scores_gemma":[0.6505179,0.0003696212,0.3441286,0.0002651733,0.0001219657,0.0001943925,0.003313276,0.0001673004,0.0009217114],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003692213,"threshold_uncertainty_score":0.01892507,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02467601805185922,"score_gpt":0.292098508945551,"score_spread":0.2674224908936918,"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."}}