{"id":"W2962752375","doi":"10.4230/lipics.icdt.2019.6","title":"A Formal Framework for Probabilistic Unclean Databases.","year":2019,"lang":"en","type":"article","venue":"DROPS (Schloss Dagstuhl – Leibniz Center for Informatics)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Probabilistic logic; Computer science; Tuple; Cardinality (data modeling); Database; Probabilistic database; Realization (probability); Data mining; Theoretical computer science; Database theory; Database design; 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":[0.01664356,0.001736876,0.001560747,0.004538955,0.002910065,0.009443833,0.007617663,0.004219567,0.009494969],"category_scores_gemma":[0.03499837,0.002361881,0.006052098,0.005791965,0.008919773,0.01963958,0.01096869,0.009946758,0.002901128],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004734007,"about_ca_system_score_gemma":0.005305792,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006224689,"about_ca_topic_score_gemma":0.004551849,"domain_scores_codex":[0.9816391,0.005825458,0.002138463,0.003962256,0.005409418,0.001025351],"domain_scores_gemma":[0.9687572,0.01917444,0.002392842,0.005818984,0.002974364,0.0008822261],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002112031,0.00002599408,0.0002551866,0.0001149257,0.00002729472,0.000131855,0.0002918835,0.009814603,0.0002905387,0.9777899,0.001532151,0.009704633],"study_design_scores_gemma":[0.00002373004,0.00002331513,0.00007767458,0.00008916252,0.00003004649,0.0002840674,0.0001021599,0.05013483,0.0005398064,0.9249571,0.02370533,0.00003292916],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0008294635,0.0005629461,0.9939103,0.001109821,0.00007170974,0.0001143903,0.0004585175,0.0002627785,0.002680094],"genre_scores_gemma":[0.1014207,0.001718685,0.8847226,0.001847651,0.0006301024,0.001321258,0.002252975,0.0003938925,0.005692112],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01664356,"threshold_uncertainty_score":0.08802056,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1331143116709655,"score_gpt":0.405443553735199,"score_spread":0.2723292420642335,"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."}}