{"id":"W4392238058","doi":"10.7717/peerj-cs.1888","title":"exKidneyBERT: a language model for kidney transplant pathology reports and the crucial role of extended vocabularies","year":2024,"lang":"en","type":"article","venue":"PeerJ Computer Science","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Vocabulary; Natural language processing; Artificial intelligence; Information extraction; Information retrieval; Language model; Domain (mathematical analysis); Field (mathematics); Pathology; Medicine; Linguistics","routes":{"ca_aff":true,"ca_fund":false,"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":[0.0006859644,0.00008395188,0.0001346753,0.00003983043,0.00009789358,0.00005153277,0.0002114214,0.00006888739,9.834527e-7],"category_scores_gemma":[0.0001741715,0.00005182964,0.00005836945,0.00009165894,0.00128346,0.000004748262,0.0001230041,0.00005446297,2.18303e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000003212745,"about_ca_system_score_gemma":0.0002309266,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001304714,"about_ca_topic_score_gemma":0.000005961197,"domain_scores_codex":[0.9991339,0.00002231481,0.0001624697,0.0003615943,0.0001352627,0.0001844401],"domain_scores_gemma":[0.9995713,0.00003809448,0.00003622002,0.0002260377,0.00004859026,0.00007973931],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002755105,0.00006938349,0.0001719847,0.0001826384,0.00006776215,0.0001012098,0.01414408,0.0002029742,0.6381007,0.007608771,0.00376258,0.3353123],"study_design_scores_gemma":[0.0007636515,0.0003192687,0.0005244365,0.00008118515,0.0000500193,0.0007048093,0.0001409752,0.8985221,0.08163239,0.008698129,0.008334966,0.000228087],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3108101,0.004409286,0.6818917,0.001951755,0.000523565,0.0002113285,0.00004722253,0.00003854558,0.0001165661],"genre_scores_gemma":[0.975688,0.00004658667,0.02356744,0.0004172666,0.0001227406,0.00002071881,0.00001265023,0.000005283952,0.0001192436],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8983191,"threshold_uncertainty_score":0.4728962,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.007218420904766878,"score_gpt":0.2577834305421606,"score_spread":0.2505650096373938,"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."}}