{"id":"W2792572948","doi":"10.1145/3186549.3186559","title":"Data Quality","year":2018,"lang":"en","type":"article","venue":"ACM SIGMOD Record","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":32,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; University of Waterloo","funders":"","keywords":"Computer science; Scope (computer science); Quality (philosophy); Data quality; Empiricism; Data science; Action (physics); Data mining; Epistemology; Programming language; Engineering","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.2377038,0.0008683082,0.002089032,0.008266796,0.005483202,0.02800883,0.006467119,0.004075619,0.02770416],"category_scores_gemma":[0.5065222,0.001169604,0.002711462,0.0123245,0.01107782,0.02551043,0.01457493,0.008119707,0.009645497],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01352693,"about_ca_system_score_gemma":0.03795461,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01036062,"about_ca_topic_score_gemma":0.006116215,"domain_scores_codex":[0.7142184,0.1348392,0.03182683,0.01847382,0.09605062,0.004591084],"domain_scores_gemma":[0.3651806,0.2313117,0.0320962,0.1864015,0.1724703,0.01253967],"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":[0.0001427814,0.000095048,0.01157377,0.001720267,0.0001928582,0.0001905924,0.002376759,0.001489982,0.0004210506,0.6249557,0.1121992,0.244642],"study_design_scores_gemma":[0.00005370618,0.0001164211,0.003337551,0.002637476,0.00006697126,0.0003463082,0.001434251,0.001699379,0.0006453721,0.3030515,0.6865367,0.0000742903],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006542586,0.01501703,0.4663347,0.3166806,0.008321189,0.002497678,0.007229406,0.002088456,0.1752884],"genre_scores_gemma":[0.2743502,0.01766036,0.5273928,0.09971249,0.009577886,0.005112445,0.0157381,0.002655252,0.04780054],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.2377038,"threshold_uncertainty_score":0.9400469,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6848192121373361,"score_gpt":0.55904053398058,"score_spread":0.1257786781567561,"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."}}