{"id":"W4415231396","doi":"10.1609/aies.v8i3.36712","title":"A Principled Approach for Data Bias Mitigation","year":2025,"lang":"en","type":"article","venue":"Proceedings of the AAAI/ACM Conference on AI Ethics and Society","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Canada Excellence Research Chairs, Government of Canada; National Science Foundation","keywords":"Measure (data warehouse); Table (database); Tuple; Key (lock); Work (physics); Data collection","routes":{"ca_aff":true,"ca_fund":true,"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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01360294,0.0001361674,0.0002664246,0.00005094034,0.0005264643,0.0005542681,0.003010353,0.0001952711,0.00001006002],"category_scores_gemma":[0.01636511,0.00008661476,0.0001311839,0.0005112547,0.0003764951,0.0004040895,0.002344468,0.0005689173,0.000001903685],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001941254,"about_ca_system_score_gemma":0.000201665,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000269268,"about_ca_topic_score_gemma":0.000005423344,"domain_scores_codex":[0.9977249,0.00005429603,0.0004911061,0.0006255506,0.0009156163,0.0001885359],"domain_scores_gemma":[0.9960428,0.001460309,0.0003850436,0.0009072135,0.001152446,0.00005212711],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002584679,0.00007322827,0.0007806822,0.0002528258,0.00005354943,7.653544e-9,0.003163463,0.000001976422,0.0004530883,0.9181263,0.0708349,0.006234145],"study_design_scores_gemma":[0.001011106,0.00009908732,0.003423114,0.0002673554,0.000113195,4.978041e-7,0.01913296,0.1236854,0.001868194,0.7569813,0.09313747,0.0002802849],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"commentary","genre_gemma":"empirical","genre_scores_codex":[0.1604616,0.000369694,0.1371003,0.548507,0.001383874,0.006370214,0.001356434,0.0002275168,0.1442234],"genre_scores_gemma":[0.9548246,0.0005219649,0.02858078,0.01019007,0.00004954759,0.0000663485,0.00005107338,0.000009432411,0.005706191],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.794363,"threshold_uncertainty_score":0.9919205,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5508003255245804,"score_gpt":0.4892577970435352,"score_spread":0.06154252848104513,"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."}}