{"id":"W2160997301","doi":"10.1109/icde.2007.369059","title":"μBE: User Guided Source Selection and Schema Mediation for Internet Scale Data Integration","year":2007,"lang":"en","type":"article","venue":"","topic":"Semantic Web and Ontologies","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Data integration; Schema (genetic algorithms); Iterative and incremental development; Data mining; User interface; The Internet; Optimization problem; Information retrieval; Algorithm; Software engineering; World Wide Web; Programming language","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006745217,0.00006902962,0.00007762943,0.00007218954,0.00004966872,0.000122613,0.0003857397,0.00005842938,0.000006453127],"category_scores_gemma":[0.0002154288,0.00005467389,0.00001246734,0.0001363057,0.00001784347,0.0007068795,0.0001865111,0.00004277864,0.000006212959],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001990784,"about_ca_system_score_gemma":0.00001720578,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000133617,"about_ca_topic_score_gemma":0.001379084,"domain_scores_codex":[0.9992771,0.00001330668,0.00017193,0.0002885943,0.0001092739,0.0001398357],"domain_scores_gemma":[0.9994102,0.000142664,0.00005619317,0.0002744223,0.00008247521,0.00003411423],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005276153,0.0001006599,0.03742671,0.00004387085,0.00003889367,8.872755e-7,0.00228528,0.00001979347,0.01790576,0.1139619,0.08650806,0.7416554],"study_design_scores_gemma":[0.0005587298,0.000109608,0.02163838,0.00001616496,0.00001128846,0.00002928866,0.0002972944,0.845238,0.1079381,0.002127687,0.02183886,0.0001966201],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09345699,0.00003024793,0.9048409,0.0006966935,0.0001945102,0.0001417491,5.791435e-7,0.000153691,0.0004846026],"genre_scores_gemma":[0.7335011,0.000005869721,0.2649814,0.0002529382,0.00009783688,0.000004691239,0.00002595311,0.00000396015,0.001126128],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8452182,"threshold_uncertainty_score":0.2229537,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05236232831897631,"score_gpt":0.3146645219986351,"score_spread":0.2623021936796588,"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."}}