{"id":"W4414359346","doi":"10.24963/ijcai.2025/572","title":"Optimal Transport on Categorical Data for Conterfactuals Using Compositional Data and Dirichlet Transport","year":2025,"lang":"en","type":"article","venue":"","topic":"Geochemistry and Geologic Mapping","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Université du Québec à Montréal","funders":"","keywords":"Categorical variable; Compositional data; Probabilistic logic; Multivariate statistics; Quantile; Dirichlet distribution; Euclidean geometry","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.03799113,0.001771402,0.003317108,0.003787908,0.00256596,0.005368175,0.004747139,0.0039961,0.006776093],"category_scores_gemma":[0.09737293,0.001712722,0.005016412,0.003420839,0.00723502,0.0100305,0.006028194,0.006260091,0.0008562886],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005686743,"about_ca_system_score_gemma":0.004438489,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009300926,"about_ca_topic_score_gemma":0.007364515,"domain_scores_codex":[0.9872643,0.008566336,0.0006161101,0.002144545,0.0009090009,0.0004997055],"domain_scores_gemma":[0.9376227,0.05180839,0.002811947,0.00503134,0.001948182,0.0007774201],"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":[0.0001857658,0.00006457074,0.00290994,0.0002063977,0.0001241486,0.0001949918,0.0006217248,0.314093,0.0006274337,0.6503785,0.0009549246,0.02963862],"study_design_scores_gemma":[0.00002170013,0.0000256773,0.0002835661,0.00004959366,0.00002175944,0.00004094914,0.00007847048,0.546141,0.0004414754,0.4517381,0.001126277,0.00003138806],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009278344,0.0001843529,0.9890358,0.0003899919,0.00003677039,0.00008925347,0.0001202311,0.0001155137,0.0007497823],"genre_scores_gemma":[0.3673162,0.0008556989,0.6236874,0.0005018554,0.0002326658,0.0009233404,0.00107418,0.0005337225,0.004874726],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.03799113,"threshold_uncertainty_score":0.2009187,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08998832565368778,"score_gpt":0.3200846941909689,"score_spread":0.2300963685372811,"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."}}