{"id":"W304693962","doi":"10.1007/978-3-319-13186-3_34","title":"Models for Distributed, Large Scale Data Cleaning","year":2014,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"McMaster University","funders":"","keywords":"Computer science; Data quality; Big data; Data mining; Dynamic data; Scale (ratio); Data stream mining; Variety (cybernetics); Distributed computing; Real-time computing; Database; 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":[0.004280759,0.001172414,0.001909342,0.0008727097,0.0008333764,0.003040024,0.004388974,0.002231752,0.007489313],"category_scores_gemma":[0.01772632,0.00109428,0.001922731,0.00200083,0.001196468,0.004669081,0.002306625,0.003977892,0.002448416],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001749177,"about_ca_system_score_gemma":0.001935411,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007022019,"about_ca_topic_score_gemma":0.008521091,"domain_scores_codex":[0.9979295,0.00090019,0.0001446445,0.0004277593,0.0004528107,0.0001451981],"domain_scores_gemma":[0.9892757,0.007953207,0.0004910466,0.001246003,0.0007967301,0.0002372922],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001123768,0.00008440427,0.0007402058,0.0002177138,0.0001518247,0.00008354655,0.0001173141,0.7448041,0.000552183,0.1773205,0.01325714,0.0625587],"study_design_scores_gemma":[0.00001353354,0.000009596249,0.00007205511,0.00001418273,0.00001651452,0.0000242223,0.00001225233,0.8671008,0.0001617476,0.1301738,0.002391319,0.000009956846],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001469605,0.0005551538,0.9955706,0.0005142321,0.00007658926,0.00003069954,0.0002430592,0.000463343,0.001076787],"genre_scores_gemma":[0.2400503,0.003305811,0.7297586,0.0008889638,0.000656189,0.001025605,0.002457333,0.0009247604,0.02093252],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007489313,"threshold_uncertainty_score":0.02505428,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1703087149308163,"score_gpt":0.3783480416925222,"score_spread":0.2080393267617059,"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."}}