{"id":"W4376130995","doi":"10.1016/j.jbi.2023.104384","title":"Deep learning to refine the identification of high-quality clinical research articles from the biomedical literature: Performance evaluation","year":2023,"lang":"en","type":"article","venue":"Journal of Biomedical Informatics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"Hamilton Health Sciences; McMaster University; Impact","funders":"","keywords":"Computer science; Identification (biology); Quality (philosophy); Data science; Artificial intelligence; Deep learning","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.02180002,0.0001060407,0.0002589516,0.0001629601,0.000210329,0.00006860776,0.0006968148,0.0002941414,0.00001939105],"category_scores_gemma":[0.0135411,0.00005379191,0.0001332436,0.001161405,0.000927526,0.00001699787,0.0002774502,0.0007636083,0.00003801194],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002542446,"about_ca_system_score_gemma":0.0002074039,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001235334,"about_ca_topic_score_gemma":0.000004412527,"domain_scores_codex":[0.9950451,0.0008579299,0.001985692,0.0001189994,0.001705599,0.0002866778],"domain_scores_gemma":[0.9968457,0.001032964,0.0007190389,0.0003815442,0.0008298503,0.0001908618],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0003640147,0.0002021212,0.00269116,0.00007110776,0.0002148905,0.000003458825,0.005076825,0.0002047765,0.02446917,0.00005883725,0.03879716,0.9278465],"study_design_scores_gemma":[0.004512803,0.006244,0.539605,0.0008526963,0.0002556493,0.00009762686,0.02164238,0.06789605,0.02365921,0.002367806,0.3323521,0.0005146214],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9869515,0.0004499973,0.004966419,0.006849655,0.0005862046,0.0001496559,0.00001827316,0.00001037061,0.00001791571],"genre_scores_gemma":[0.9960288,0.0007666968,0.001816728,0.0003307695,0.0007905053,0.000008868952,0.0001990536,0.000007776664,0.0000507423],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9273319,"threshold_uncertainty_score":0.9947683,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.12669730255499,"score_gpt":0.452121140548754,"score_spread":0.3254238379937641,"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."}}