{"id":"W2802105481","doi":"10.18653/v1/s18-2005","title":"Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems","year":2018,"lang":"en","type":"article","venue":"","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":395,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Computer science; Task (project management); Benchmark (surveying); Race (biology); Sentiment analysis; Artificial intelligence; Affect (linguistics); Gender bias; Natural language processing; Machine learning; SemEval; Psychology; Social psychology; Sociology","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":[0.0004848949,0.00007644096,0.0001438497,0.0002932192,0.00007686699,0.0002223419,0.0001628188,0.00002536499,0.00001974815],"category_scores_gemma":[0.00001234945,0.00007150749,0.00002514913,0.0009536541,0.00002442976,0.000215182,0.0001087999,0.00005070337,0.00005641973],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003024319,"about_ca_system_score_gemma":0.00001390351,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005927914,"about_ca_topic_score_gemma":0.0001796666,"domain_scores_codex":[0.9990475,0.00008715855,0.0001725544,0.0003250185,0.0001725228,0.0001952634],"domain_scores_gemma":[0.9995095,0.00003688143,0.00004860832,0.0002968799,0.00004207837,0.00006607662],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0000603299,0.0004292618,0.6550287,0.000118306,0.002237919,0.0002475207,0.01958985,0.007056232,0.041651,0.05704163,0.001515024,0.2150242],"study_design_scores_gemma":[0.0007651875,0.000134346,0.1364999,0.000019291,0.00007836914,0.00003456173,0.0007355461,0.8494418,0.01064658,0.0001984386,0.001101729,0.000344243],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6577668,0.00007514289,0.3362503,0.00002683804,0.0002194884,0.00008147764,1.65813e-7,0.00009317388,0.005486585],"genre_scores_gemma":[0.9911086,0.000004574952,0.007828357,0.00005631112,0.00005441748,0.00000585541,5.074667e-7,0.00000337903,0.0009379956],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8423856,"threshold_uncertainty_score":0.291599,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06497271295094052,"score_gpt":0.290410277792946,"score_spread":0.2254375648420055,"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."}}