{"id":"W2189052568","doi":"10.14778/2850583.2850586","title":"The iBench integration metadata generator","year":2015,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Metadata; Computer science; Schema evolution; Data integration; Generality; Schema (genetic algorithms); Generator (circuit theory); Data mapping; Data element; Data science; Information retrieval; Data mining; Database; World Wide Web; Database schema","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":[0.0105629,0.001574923,0.0009364938,0.004237203,0.00121775,0.003857498,0.004111745,0.001412171,0.01043958],"category_scores_gemma":[0.03922306,0.001501764,0.001345227,0.0031702,0.00121733,0.005270116,0.007125597,0.003041398,0.005484642],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001539215,"about_ca_system_score_gemma":0.003282546,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003205616,"about_ca_topic_score_gemma":0.002250475,"domain_scores_codex":[0.9949244,0.001272997,0.0006846364,0.0006499757,0.002138867,0.0003291011],"domain_scores_gemma":[0.981302,0.006048956,0.0007543368,0.007343572,0.003713557,0.0008376914],"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":[0.00358912,0.002041537,0.03016826,0.002539309,0.0004542899,0.001998286,0.003623036,0.07137456,0.03058759,0.08814289,0.2579482,0.507533],"study_design_scores_gemma":[0.001094221,0.0008444972,0.005945023,0.0004959655,0.0001933744,0.0009596358,0.0009220625,0.5072443,0.07644674,0.06621771,0.3392588,0.0003777639],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.04572805,0.0006375628,0.4935828,0.001139347,0.0006122246,0.002682892,0.0154248,0.4169441,0.02324813],"genre_scores_gemma":[0.2286062,0.0006011627,0.6572211,0.0009384357,0.0001508678,0.003694789,0.05728608,0.03996333,0.01153793],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.0105629,"threshold_uncertainty_score":0.05586261,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2684436286050799,"score_gpt":0.3942632791998347,"score_spread":0.1258196505947548,"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."}}