{"id":"W3101628588","doi":"10.1371/journal.pone.0241239","title":"A machine learning approach to predict ethnicity using personal name and census location in Canada","year":2020,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Names, Identity, and Discrimination Research","field":"Social Sciences","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Canadian Institutes of Health Research; University of Alberta; Government of Canada; Alberta Machine Intelligence Institute","keywords":"Census; Ethnic group; Artificial intelligence; Substring; Support vector machine; Population; Machine learning; Computer science; Logistic regression; Demography; Set (abstract data type); Sociology; Political science; Law","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.001116087,0.0006279114,0.0004596212,0.002909745,0.001842176,0.001332738,0.001135189,0.0004079589,0.001759018],"category_scores_gemma":[0.004285761,0.0002128336,0.0007308564,0.002671601,0.0004554804,0.0004579996,0.0009663721,0.0009819256,0.0005208235],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01179054,"about_ca_system_score_gemma":0.02019968,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.9636421,"about_ca_topic_score_gemma":0.9572978,"domain_scores_codex":[0.9994701,0.00007992663,0.00002930105,0.0001294655,0.0001656216,0.0001256985],"domain_scores_gemma":[0.9986512,0.0003127198,0.0001068138,0.00005991777,0.0007551485,0.0001143233],"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":[0.0002070496,0.0002025722,0.7712334,0.00008697629,0.0002007456,0.0003661124,0.0005474543,0.06036875,0.000947903,0.002071112,0.01026889,0.153499],"study_design_scores_gemma":[0.00003037756,0.00004572065,0.2119663,0.00006770323,0.0001155592,0.0001399475,0.001366015,0.776334,0.001509058,0.002349082,0.006017445,0.00005881507],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9198404,0.0009164633,0.05418323,0.002200042,0.00008123215,0.0003352098,0.01161139,0.001145431,0.009686622],"genre_scores_gemma":[0.9590003,0.0002643658,0.03292131,0.0001283387,0.00001750024,0.00007225126,0.00459469,0.00003753028,0.002963752],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03635788,"threshold_uncertainty_score":0.08554679,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1537226894422716,"score_gpt":0.3131259734773718,"score_spread":0.1594032840351002,"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."}}