{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.06773838,0.001929928,0.003021099,0.005872388,0.002770229,0.00764264,0.008625802,0.004703133,0.004246944],"category_scores_gemma":[0.1568659,0.001791525,0.005270676,0.00600155,0.008458903,0.009903845,0.01976114,0.01036042,0.002037068],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002229373,"about_ca_system_score_gemma":0.008800176,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008747202,"about_ca_topic_score_gemma":0.0008446238,"domain_scores_codex":[0.9151098,0.03957108,0.003983206,0.01040145,0.02935263,0.001581913],"domain_scores_gemma":[0.8470774,0.08795191,0.006171804,0.04235829,0.01478549,0.001655187],"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":[0.0002521862,0.0002134764,0.003933004,0.0006710691,0.0006053455,0.0002548464,0.001100509,0.03185279,0.003250096,0.6763695,0.01101259,0.2704844],"study_design_scores_gemma":[0.0001000633,0.0001118037,0.0004196555,0.0001957339,0.0001124201,0.0003669478,0.000111575,0.1074235,0.003102645,0.869004,0.01898374,0.000067883],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0006902806,0.0001879716,0.9963275,0.001282218,0.00008619468,0.0001358743,0.00009595467,0.0002683759,0.0009256958],"genre_scores_gemma":[0.05417547,0.0003740257,0.9401846,0.001744378,0.0005927387,0.001029207,0.0003748883,0.000188893,0.001335881],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.06773838,"threshold_uncertainty_score":0.3582391,"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."}}