{"id":"W4378232119","doi":"10.1093/jssam/smad015","title":"Automated Classification for Open-Ended Questions with BERT","year":2023,"lang":"en","type":"article","venue":"Journal of Survey Statistics and Methodology","topic":"Topic Modeling","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; Western University","funders":"Social Sciences and Humanities Research Council; Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Coding (social sciences); Artificial intelligence; Boosting (machine learning); Natural language processing; Machine learning; Language model; Training set; Statistics","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":[0.009821454,0.0009717103,0.0007556367,0.002442108,0.0004476257,0.001281019,0.001521868,0.001466596,0.004296798],"category_scores_gemma":[0.05542549,0.0003372126,0.0006503614,0.001565665,0.0006401632,0.002610463,0.001649653,0.002226998,0.003621974],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00120911,"about_ca_system_score_gemma":0.0009440864,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002051675,"about_ca_topic_score_gemma":0.002885373,"domain_scores_codex":[0.9930335,0.004431301,0.0003845817,0.0009350243,0.0008485885,0.0003670184],"domain_scores_gemma":[0.9385139,0.04560056,0.00287909,0.006273583,0.005735077,0.0009978196],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001980311,0.001055633,0.09056731,0.0009987942,0.0001489321,0.000266471,0.002277793,0.06424219,0.01685926,0.009624635,0.04375131,0.7682275],"study_design_scores_gemma":[0.00006830709,0.0003063315,0.02997775,0.0001200293,0.00003128884,0.0001675511,0.0007839292,0.9298141,0.01379093,0.01450488,0.01037114,0.00006373871],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5554429,0.0009808594,0.4126046,0.001689665,0.0003507421,0.0008010605,0.005922379,0.01389659,0.008311239],"genre_scores_gemma":[0.8951675,0.00009474348,0.09531221,0.0002027122,0.00007845092,0.0003758139,0.006134529,0.0001856511,0.002448325],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009821454,"threshold_uncertainty_score":0.05194145,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4858752837440197,"score_gpt":0.4643715025409455,"score_spread":0.02150378120307417,"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."}}