{"id":"W4391096379","doi":"10.1109/bigdata59044.2023.10386531","title":"Refactoring ETL Flows in The Wild","year":2023,"lang":"en","type":"article","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Code refactoring; Computer science; IBM; Context (archaeology); Pipeline (software); Data integration; Data flow diagram; Data mining; Database; Software engineering; Operating system; Software","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":[0.0002580715,0.00003128845,0.00002958908,0.00003895417,0.00005579007,0.00009992011,0.0007209381,0.00001107513,0.000005089307],"category_scores_gemma":[0.00001230341,0.0000195671,0.00001065793,0.0006941846,0.000004883871,0.0001620455,0.0001300654,0.00005570748,0.0003080231],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000005424266,"about_ca_system_score_gemma":0.000009285448,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008129574,"about_ca_topic_score_gemma":0.00003365077,"domain_scores_codex":[0.9995897,0.00001163103,0.00006325332,0.0001291036,0.00009633222,0.0001099778],"domain_scores_gemma":[0.9995037,0.00006915655,0.000008761616,0.0003988321,0.000005569823,0.00001396717],"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":[1.931495e-7,0.00003285213,0.0008258044,0.00000291047,0.000002849999,0.00001323905,0.002615003,0.0001133692,0.0002576905,0.2516386,0.03394703,0.7105504],"study_design_scores_gemma":[0.0001162449,0.00001594495,0.03290016,0.00000955435,9.109408e-7,0.000005555481,0.0003499131,0.697019,0.0001800522,0.006968092,0.262304,0.0001305703],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1863299,0.00003180533,0.7211845,0.02861158,0.0005523579,0.0003340817,0.000009800383,0.001188191,0.06175774],"genre_scores_gemma":[0.8878474,0.00002835566,0.1088823,0.001072254,0.0001352237,0.0001006436,0.00001319617,0.000006779853,0.001913942],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7104198,"threshold_uncertainty_score":0.3959117,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04099050327581285,"score_gpt":0.2888281117088855,"score_spread":0.2478376084330726,"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."}}