{"id":"W2021739502","doi":"10.5555/1182635.1164157","title":"Multi-column substring matching for database schema translation","year":2006,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":41,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Substring; Computer science; Database schema; Schema (genetic algorithms); Schema matching; Data mining; Algorithm; Artificial intelligence; Set (abstract data type); Database; Theoretical computer science; Database design; Information retrieval; Programming language; Data integration","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":[0.0002160658,0.00008592955,0.00007908184,0.00007160473,0.0001091718,0.0001695871,0.0004224314,0.00004337467,0.000003918526],"category_scores_gemma":[0.00001442248,0.0000781441,0.00003853526,0.0001765221,0.00001035645,0.0008648394,0.00005689987,0.000074991,0.000003666654],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002213139,"about_ca_system_score_gemma":0.00001909114,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002326023,"about_ca_topic_score_gemma":0.0001163285,"domain_scores_codex":[0.9992784,0.0000102887,0.000153161,0.0002571058,0.0001203564,0.0001806505],"domain_scores_gemma":[0.9995552,0.00006685621,0.00004739179,0.0002577621,0.00004845537,0.00002434165],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00001126213,0.0001360928,0.0002600794,0.0001592836,0.000008040595,0.00001250931,0.0003523541,0.00005580992,0.67927,0.1806468,0.00119431,0.1378935],"study_design_scores_gemma":[0.0007523549,0.00002339766,0.0001359595,0.00007490131,0.000007142756,0.00001438274,0.00001477319,0.2618476,0.6903553,0.04530875,0.001101152,0.0003642034],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004806856,0.0006839552,0.9928237,0.0002943703,0.00005741327,0.0002214893,0.000003707284,0.0009115451,0.0001969169],"genre_scores_gemma":[0.2964906,8.003794e-7,0.7032182,0.00006371497,0.0000358984,0.00001830373,0.00001343496,0.000006078763,0.0001529768],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2916838,"threshold_uncertainty_score":0.3186624,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03110730105148172,"score_gpt":0.2964350817566067,"score_spread":0.2653277807051249,"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."}}