{"id":"W3097720724","doi":"10.2196/19612","title":"Amplifying Domain Expertise in Clinical Data Pipelines","year":2020,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Institute of Allergy and Infectious Diseases; National Institutes of Health; National Science Foundation","keywords":"Computer science; Domain (mathematical analysis); Data science; Pipeline (software); USable; Automatic summarization; Subject-matter expert; Data curation; Visualization; Process (computing); Domain model; Domain knowledge; Software engineering; Data mining; Artificial intelligence; Expert system; World Wide Web","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":[0.05825445,0.001613001,0.001180701,0.01035932,0.003483175,0.008449589,0.004035039,0.003634914,0.004766524],"category_scores_gemma":[0.1514732,0.001886817,0.001933872,0.005691984,0.006974316,0.01843445,0.02194723,0.004966414,0.001854318],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004518239,"about_ca_system_score_gemma":0.006087009,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003517629,"about_ca_topic_score_gemma":0.003086357,"domain_scores_codex":[0.9369583,0.03309887,0.005555181,0.006982994,0.01424458,0.003160077],"domain_scores_gemma":[0.7479872,0.1849002,0.01581094,0.0235689,0.02337879,0.004353872],"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":[0.0008481276,0.000505719,0.02987534,0.004553938,0.0002590991,0.001554012,0.07840651,0.01269785,0.0189977,0.05853224,0.01765733,0.7761122],"study_design_scores_gemma":[0.0004541611,0.001303493,0.02973136,0.004126372,0.0007618057,0.007250262,0.02874372,0.1464812,0.05420835,0.3411371,0.3850064,0.0007957393],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.09396479,0.003530604,0.8627009,0.01351613,0.0001512175,0.001391586,0.0003318276,0.006289167,0.01812366],"genre_scores_gemma":[0.433618,0.001424912,0.5555289,0.002384363,0.0002254635,0.0008415185,0.0008131067,0.000649183,0.004514531],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.05825445,"threshold_uncertainty_score":0.3080826,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4335152015891584,"score_gpt":0.5153609075619537,"score_spread":0.08184570597279534,"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."}}