{"id":"W2157044362","doi":"10.1007/s00778-008-0128-8","title":"Top-k typicality queries and efficient query answering methods on large databases","year":2009,"lang":"en","type":"article","venue":"The VLDB Journal","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":false,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Computer science; Discriminative model; Set (abstract data type); Tree (set theory); Information retrieval; Measure (data warehouse); Quality (philosophy); Series (stratigraphy); 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":[0.008500651,0.001527301,0.004818978,0.00334122,0.0024818,0.006715759,0.007021165,0.003233871,0.005305656],"category_scores_gemma":[0.04078459,0.002013889,0.002561387,0.009267921,0.003048939,0.01916537,0.005584576,0.003811864,0.002040759],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002216254,"about_ca_system_score_gemma":0.002344378,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00274866,"about_ca_topic_score_gemma":0.003151962,"domain_scores_codex":[0.9869449,0.003939845,0.001567791,0.002749191,0.003823037,0.0009751743],"domain_scores_gemma":[0.964893,0.02176647,0.00140174,0.008451535,0.002760776,0.0007264594],"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":[0.004140123,0.0004978483,0.00950204,0.002076219,0.0006286331,0.0005512723,0.001694934,0.1967379,0.01605605,0.2537173,0.05391936,0.4604784],"study_design_scores_gemma":[0.000147047,0.0001177571,0.0005880591,0.00004036884,0.0001389733,0.0005980825,0.0003120322,0.5930769,0.004404963,0.3957084,0.0048062,0.00006128094],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0323919,0.002384635,0.9565058,0.001988192,0.0001420678,0.0001011085,0.001177637,0.003597562,0.001711157],"genre_scores_gemma":[0.4254284,0.001441173,0.563081,0.0007538561,0.0007796686,0.0002725379,0.003677942,0.001061245,0.003504214],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008500651,"threshold_uncertainty_score":0.04495627,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03418460099070763,"score_gpt":0.3495786809841878,"score_spread":0.3153940799934802,"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."}}