{"id":"W2435238507","doi":"","title":"Data Profiling","year":2018,"lang":"en","type":"book","venue":"","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Profiling (computer programming); Computer science; Metadata; Inference; Data mining; Data science; Artificial intelligence; World Wide Web; Programming language","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":[0.006754078,0.00129423,0.001325898,0.005425397,0.001532019,0.007729754,0.002791533,0.001485586,0.01900609],"category_scores_gemma":[0.02811282,0.0009781235,0.001451452,0.01147618,0.0009649429,0.01403164,0.005459277,0.002988289,0.02018362],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00167802,"about_ca_system_score_gemma":0.002398788,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00102673,"about_ca_topic_score_gemma":0.001403184,"domain_scores_codex":[0.9933114,0.001360997,0.0005637496,0.001251421,0.003239556,0.0002729815],"domain_scores_gemma":[0.9817349,0.00692097,0.0006880453,0.006735859,0.003378011,0.0005421739],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001091503,0.00007353588,0.001949702,0.001336792,0.00005654136,0.0002093247,0.000660955,0.003090158,0.004142718,0.1746769,0.1780031,0.6356911],"study_design_scores_gemma":[0.000009398,0.00003099969,0.0008847088,0.0006637423,0.00002818065,0.0006660806,0.0002253836,0.006016966,0.004960143,0.08951241,0.8969573,0.00004461824],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"other","genre_scores_codex":[0.00387471,0.0219276,0.8349635,0.007836225,0.001695863,0.0009823502,0.01884035,0.01264372,0.09723566],"genre_scores_gemma":[0.06282921,0.03822896,0.7440152,0.006115002,0.002264297,0.001354381,0.05380961,0.007045896,0.08433737],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.01900609,"threshold_uncertainty_score":0.06358176,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6744232901337949,"score_gpt":0.5236971452500062,"score_spread":0.1507261448837887,"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."}}