{"id":"W4224646058","doi":"10.18653/v1/2022.bionlp-1.32","title":"ICDBigBird: A Contextual Embedding Model for ICD Code Classification","year":2022,"lang":"en","type":"article","venue":"","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":23,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; University of Toronto","funders":"Mitacs","keywords":"Computer science; Word embedding; Task (project management); Embedding; Process (computing); Artificial intelligence; Health care; ICD-10; Machine learning; Graph; Natural language processing; Theoretical computer science","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":[0.001363534,0.0009513447,0.0007769339,0.001847154,0.0006374038,0.001378165,0.001613339,0.001018729,0.006229538],"category_scores_gemma":[0.006033785,0.0003877701,0.001445593,0.001737892,0.0003861735,0.002069796,0.002463117,0.00190284,0.003569027],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009020293,"about_ca_system_score_gemma":0.001525917,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01579463,"about_ca_topic_score_gemma":0.02982164,"domain_scores_codex":[0.9989997,0.0003083069,0.00009117975,0.0002805659,0.0002455918,0.00007469071],"domain_scores_gemma":[0.9988482,0.0004890296,0.00007133034,0.0002627501,0.0002681684,0.00006055582],"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":[0.0007816409,0.0004258118,0.009938711,0.001050292,0.000492244,0.0004378649,0.0005426535,0.06114379,0.004050682,0.02726311,0.1380491,0.7558241],"study_design_scores_gemma":[0.000129259,0.0001809742,0.003862906,0.0003067422,0.0002142014,0.000393287,0.0003139001,0.7842469,0.005366744,0.09076396,0.1141122,0.000108929],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01510394,0.001955037,0.9246132,0.001355866,0.0005653553,0.000485106,0.03355199,0.01821715,0.004152359],"genre_scores_gemma":[0.1657187,0.001532433,0.7414713,0.001138256,0.0002247828,0.0009759378,0.08032516,0.001515579,0.007097811],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01579463,"threshold_uncertainty_score":0.03140539,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06264871263909677,"score_gpt":0.3368751190488523,"score_spread":0.2742264064097555,"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."}}