{"id":"W4293582904","doi":"10.1561/1900000045","title":"Trends in Cleaning Relational Data: Consistency and Deduplication","year":2015,"lang":"en","type":"article","venue":"Foundations and Trends in Databases","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":124,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Data deduplication; Consistency (knowledge bases); Data consistency; Computer science; Database; Data science; Artificial intelligence","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":[0.01936368,0.0008981948,0.001448277,0.005585612,0.001948335,0.008655013,0.006440555,0.002983561,0.003066605],"category_scores_gemma":[0.05149089,0.001273606,0.001486957,0.01406191,0.00411055,0.02428659,0.00617783,0.005761421,0.001714729],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003144025,"about_ca_system_score_gemma":0.003278563,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002630088,"about_ca_topic_score_gemma":0.001844307,"domain_scores_codex":[0.9747591,0.005694814,0.001904637,0.003545117,0.01334986,0.0007464622],"domain_scores_gemma":[0.9262832,0.03497101,0.005502846,0.01700852,0.01502786,0.001206605],"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":[0.0002429244,0.0001803526,0.01311288,0.003412206,0.0001574203,0.000290401,0.001638449,0.008498482,0.003979054,0.249065,0.05315571,0.6662672],"study_design_scores_gemma":[0.00006272256,0.000258264,0.00901033,0.002193359,0.0001604195,0.003379347,0.002766653,0.05077135,0.01655716,0.262406,0.6522445,0.0001899801],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03055187,0.3145616,0.5577621,0.06270677,0.003379966,0.0004419772,0.001870413,0.002319751,0.02640553],"genre_scores_gemma":[0.2343031,0.254038,0.4696408,0.01610739,0.008358362,0.0005310237,0.006538611,0.001286358,0.009196432],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01936368,"threshold_uncertainty_score":0.1024062,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2018429451686717,"score_gpt":0.3689833687587856,"score_spread":0.1671404235901139,"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."}}