{"id":"W2250194349","doi":"10.18653/v1/d13-1114","title":"Gender Inference of Twitter Users in Non-English Contexts","year":2013,"lang":"en","type":"article","venue":"","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":133,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada; Public Safety Canada","keywords":"Inference; Computer science; Orthography; Natural language processing; Social media; Artificial intelligence; Data science; Linguistics; World Wide Web","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":[0.003245334,0.0003673969,0.000418167,0.001139047,0.0006124938,0.001409627,0.0004517835,0.0005741806,0.004013023],"category_scores_gemma":[0.02160423,0.000197603,0.0004769192,0.001113663,0.0004001763,0.002706019,0.00106195,0.0008982324,0.001693227],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004454716,"about_ca_system_score_gemma":0.0003816479,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006907291,"about_ca_topic_score_gemma":0.009454398,"domain_scores_codex":[0.9980797,0.0009908347,0.0001183961,0.0003470621,0.0002200308,0.0002440581],"domain_scores_gemma":[0.9915684,0.005205333,0.00086044,0.0009960046,0.001057472,0.00031246],"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":[0.001167914,0.0002032788,0.8532783,0.0001609648,0.000148225,0.0004031429,0.006216015,0.004010012,0.004864778,0.008059033,0.006013799,0.1154745],"study_design_scores_gemma":[0.0000730692,0.0002584507,0.6230617,0.0002628474,0.000234982,0.0009503006,0.01884604,0.2667872,0.02074713,0.04776133,0.02083877,0.0001781422],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9428589,0.0004909392,0.03957625,0.001415924,0.000107107,0.00004805966,0.004320519,0.0002132812,0.0109691],"genre_scores_gemma":[0.9929678,0.0001117557,0.004011413,0.0000864138,0.00004031244,0.00001342173,0.00143553,0.000023853,0.001309394],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006907291,"threshold_uncertainty_score":0.01716322,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04556533857021437,"score_gpt":0.2934247514939073,"score_spread":0.2478594129236929,"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."}}