{"id":"W2981443541","doi":"10.2196/14782","title":"Efficient Reuse of Natural Language Processing Models for Phenotype-Mention Identification in Free-text Electronic Medical Records: A Phenotype Embedding Approach","year":2019,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"King's College London; Medical Research Council; Engineering and Physical Sciences Research Council; National Institute for Health and Care Research; Department of Health and Social Care; South London and Maudsley NHS Foundation Trust","keywords":"Computer science; Retraining; Identification (biology); Artificial intelligence; Machine learning; Reuse; Natural language processing; Language model; Adaptation (eye); Embedding","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":[0.005325323,0.001405145,0.001052845,0.002637306,0.0007233525,0.001927247,0.002829162,0.001570439,0.001216258],"category_scores_gemma":[0.02100939,0.0009161343,0.002269213,0.001801222,0.0009892578,0.004798238,0.003598474,0.002398779,0.001195119],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001064966,"about_ca_system_score_gemma":0.002120233,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005655605,"about_ca_topic_score_gemma":0.006976149,"domain_scores_codex":[0.9958172,0.001950585,0.0003658065,0.001090328,0.0005974082,0.000178745],"domain_scores_gemma":[0.9828463,0.008723538,0.001389802,0.005050078,0.001689751,0.0003005172],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0005388051,0.0009714155,0.01764308,0.0004706857,0.0003963917,0.00119333,0.001996907,0.2456882,0.01935323,0.009305909,0.007867862,0.6945742],"study_design_scores_gemma":[0.00002998312,0.0001175483,0.001578781,0.0000424301,0.0001232078,0.0003042004,0.0002892275,0.9693527,0.009400275,0.01463246,0.004078978,0.00005019964],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04696528,0.0002738362,0.943867,0.0007460085,0.00005135954,0.000230668,0.0005241357,0.006651955,0.0006898694],"genre_scores_gemma":[0.3752406,0.0003059429,0.6160921,0.0004885516,0.00008180509,0.0003394632,0.004248803,0.0007796567,0.00242311],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005655605,"threshold_uncertainty_score":0.02816331,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0110460746108385,"score_gpt":0.3162595550492472,"score_spread":0.3052134804384087,"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."}}