{"id":"W2950021574","doi":"10.1093/bioinformatics/btz504","title":"Towards reliable named entity recognition in the biomedical domain","year":2019,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Topic Modeling","field":"Computer Science","cited_by":88,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"National Human Genome Research Institute; Compute Canada; National Institutes of Health; Nvidia","keywords":"CRFS; Conditional random field; Computer science; Artificial intelligence; Dropout (neural networks); Transfer of learning; Named-entity recognition; Natural language processing; Machine learning; Task (project management); Regularization (linguistics); Deep learning; Sequence labeling; Source code; Generalization; Multi-task learning","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008242377,0.00007064125,0.00009284179,0.00008820405,0.00003828472,0.0001216438,0.0006759341,0.00006562908,0.00002965608],"category_scores_gemma":[0.00004261735,0.00004830406,0.00003277446,0.000313805,0.00002128975,0.0005405624,0.0001371666,0.0001407324,0.0005586942],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004166329,"about_ca_system_score_gemma":0.00007126842,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003690306,"about_ca_topic_score_gemma":0.00000520023,"domain_scores_codex":[0.9990139,0.00002888057,0.0002993145,0.0000954182,0.0003617691,0.0002007385],"domain_scores_gemma":[0.9993755,0.00004071353,0.00007123245,0.0004484148,0.00002730845,0.00003680751],"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":[0.0000188998,0.0003969847,0.003796571,0.0005945372,0.00003232533,0.00002730974,0.0517214,0.0002234101,0.0002166151,0.096962,0.01379324,0.8322167],"study_design_scores_gemma":[0.000760508,0.00009149227,0.001334302,0.00007985529,0.000003076007,0.00003120474,0.001046056,0.9359299,0.0001212581,0.02467263,0.03571289,0.0002168309],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2498937,0.00002029547,0.7249575,0.001856943,0.0007894712,0.0003726174,0.000004229385,0.00009502903,0.02201014],"genre_scores_gemma":[0.3913306,0.00001946902,0.606612,0.001798288,0.00007719224,0.00001804333,0.0000261934,0.000005124849,0.0001130759],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9357065,"threshold_uncertainty_score":0.718107,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02171330360008556,"score_gpt":0.2395539452077731,"score_spread":0.2178406416076875,"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."}}