{"id":"W7134954916","doi":"10.1109/icdmw69685.2025.00343","title":"Enhancing LLM Fine-Tuning for Text-to-SQLs by SQL Quality Measurement","year":2025,"lang":"","type":"article","venue":"","topic":"SAS software applications and methods","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Quality (philosophy); SQL; Measure (data warehouse); Data collection; Work (physics); Data quality","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004102638,0.00189838,0.0009171609,0.001556246,0.0003740282,0.002382276,0.002917399,0.000957081,0.004031616],"category_scores_gemma":[0.02854824,0.0005878499,0.001257736,0.0008838426,0.0007009188,0.003545481,0.00220762,0.002028278,0.00286914],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00116187,"about_ca_system_score_gemma":0.001997651,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003658459,"about_ca_topic_score_gemma":0.004272772,"domain_scores_codex":[0.9953166,0.001190468,0.0006956654,0.000942936,0.001602221,0.0002520975],"domain_scores_gemma":[0.9878912,0.00523067,0.0009685391,0.003263092,0.00224948,0.0003971129],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001892534,0.001608921,0.03332042,0.001256344,0.0004254095,0.0004968185,0.0007041259,0.1335838,0.1000119,0.004916381,0.02610824,0.6956752],"study_design_scores_gemma":[0.00008421863,0.0002656665,0.002376905,0.000038803,0.0000574434,0.0001222555,0.0001419856,0.9410424,0.04835641,0.002597351,0.004874357,0.00004218212],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1149413,0.0005117794,0.7024691,0.0005074459,0.0001824168,0.0006477576,0.001561624,0.1767158,0.002462811],"genre_scores_gemma":[0.5762193,0.0002139815,0.4092111,0.0005246063,0.00006258565,0.0005314471,0.005586261,0.005153043,0.002497727],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004102638,"threshold_uncertainty_score":0.02169704,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05747556567261453,"score_gpt":0.3639704449943609,"score_spread":0.3064948793217464,"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."}}