{"id":"W4403888184","doi":"10.2196/52897","title":"Multifaceted Natural Language Processing Task–Based Evaluation of Bidirectional Encoder Representations From Transformers Models for Bilingual (Korean and English) Clinical Notes: Algorithm Development and Validation","year":2024,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Security token; Encoder; Transformer; Natural language processing; Artificial intelligence; Language model; Context (archaeology); Machine learning; Engineering","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00375327,0.002442234,0.001042111,0.001426454,0.0005611246,0.001142962,0.001680783,0.001515185,0.00315611],"category_scores_gemma":[0.009712236,0.0005057554,0.001235642,0.0008268006,0.0004336175,0.001474824,0.001450205,0.002547904,0.001160976],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001669972,"about_ca_system_score_gemma":0.002364881,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01813596,"about_ca_topic_score_gemma":0.01992391,"domain_scores_codex":[0.9990269,0.0003743966,0.0001037006,0.0002709314,0.0001122527,0.0001117076],"domain_scores_gemma":[0.9949391,0.003641404,0.0001871359,0.0002761037,0.0007890045,0.0001672332],"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":[0.001848088,0.001361655,0.02016117,0.0007566686,0.0006956367,0.0006305001,0.0005270839,0.3103844,0.009072952,0.001126945,0.01286625,0.6405687],"study_design_scores_gemma":[0.0001152887,0.0002910446,0.002073766,0.00004018502,0.00009798137,0.0001051147,0.0001790357,0.9916663,0.003866608,0.0008519844,0.0006870282,0.00002574354],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.7324501,0.004174966,0.2389443,0.001147386,0.0004558674,0.001139371,0.004050943,0.01331935,0.004317719],"genre_scores_gemma":[0.8639966,0.000777562,0.1206284,0.0003668266,0.00008416968,0.0008123097,0.01032569,0.0002958371,0.002712581],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.01813596,"threshold_uncertainty_score":0.03606081,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05428755884584214,"score_gpt":0.4284199307430597,"score_spread":0.3741323718972175,"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."}}