{"id":"W2109651326","doi":"10.2196/medinform.4321","title":"Benchmarking Clinical Speech Recognition and Information Extraction: New Data, Methods, and Evaluations","year":2015,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Hospital Admissions and Outcomes","field":"Medicine","cited_by":57,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Institute of General Medical Sciences; Australian Research Council; National Institutes of Health; University of Queensland; Queensland University of Technology; Griffith University; University of Melbourne; Monash University; Australian Government; University of Sydney; Australian National University; University of New South Wales; ACT Government","keywords":"Computer science; Handover; Workflow; Benchmarking; Natural language processing; Information extraction; Speech recognition; Correctness; Data extraction; Artificial intelligence; Information retrieval; Database; MEDLINE","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002866978,0.0001182898,0.0002798436,0.0001185636,0.00007259018,0.00007145394,0.0000965364,0.0002384402,0.0003433078],"category_scores_gemma":[0.003711399,0.00008827007,0.00003093941,0.0001704286,0.0001047519,0.002221785,0.000205979,0.0004302902,0.0001107156],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002990341,"about_ca_system_score_gemma":0.0006281575,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002061973,"about_ca_topic_score_gemma":0.000003724976,"domain_scores_codex":[0.9980627,0.00007565941,0.0009388645,0.000095398,0.0006675912,0.0001597333],"domain_scores_gemma":[0.9978076,0.0003393712,0.0002092297,0.000314363,0.000177774,0.001151673],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00003011708,0.00003849576,0.002952474,0.00007843726,0.0000235002,0.000003157927,0.001270394,9.029868e-8,6.303608e-7,0.00004333171,0.02146291,0.9740965],"study_design_scores_gemma":[0.008002785,0.00126521,0.07713382,0.0007979198,0.0003951771,0.001580555,0.008977629,0.1055008,0.00006016457,0.002888707,0.7929048,0.0004923868],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.8386517,0.0007188355,0.1044473,0.01597113,0.002111782,0.002304216,0.00007284378,0.0003389475,0.03538321],"genre_scores_gemma":[0.02357532,0.00163537,0.962095,0.008913426,0.001339055,0.0000276042,0.001985706,0.00001898968,0.0004095543],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9736041,"threshold_uncertainty_score":0.4443159,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2116328860060116,"score_gpt":0.5129666329524059,"score_spread":0.3013337469463943,"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."}}