{"id":"W760598031","doi":"10.1007/s00778-015-0389-y","title":"Profiling relational data: a survey","year":2015,"lang":"en","type":"article","venue":"The VLDB Journal","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":273,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Profiling (computer programming); Metadata; Computer science; Data mining; Data type; Information retrieval; Data science; World Wide Web","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":[0.01869769,0.0006774935,0.001507291,0.01501409,0.0009242385,0.007538349,0.002744373,0.001273704,0.001958959],"category_scores_gemma":[0.07480663,0.0007422493,0.001051805,0.0222602,0.001718179,0.01735284,0.003062729,0.001598523,0.001021985],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001464438,"about_ca_system_score_gemma":0.002510754,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002434847,"about_ca_topic_score_gemma":0.003064253,"domain_scores_codex":[0.98079,0.003948452,0.002594439,0.002924362,0.009087564,0.0006552319],"domain_scores_gemma":[0.8604422,0.09986163,0.007603403,0.01009019,0.01982508,0.00217755],"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":[0.0001678369,0.0001770685,0.0908149,0.00369295,0.0001898058,0.00007622846,0.001507676,0.001936789,0.00225872,0.01415405,0.01142764,0.8735963],"study_design_scores_gemma":[0.000067401,0.0007884558,0.1373759,0.009211984,0.0006429999,0.003835496,0.01551866,0.03142736,0.01908045,0.06174427,0.7199531,0.0003539018],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.204822,0.4718415,0.2501662,0.02193191,0.0006169153,0.0005095127,0.0106903,0.001940897,0.03748073],"genre_scores_gemma":[0.4661339,0.3631345,0.1421929,0.004883834,0.0009596304,0.0004179132,0.01806351,0.0006093284,0.003604455],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.01869769,"threshold_uncertainty_score":0.09888405,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8844173631534311,"score_gpt":0.5434835487839118,"score_spread":0.3409338143695193,"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."}}