{"id":"W3199267631","doi":"10.2196/29398","title":"A Deep Learning Approach to Refine the Identification of High-Quality Clinical Research Articles From the Biomedical Literature: Protocol for Algorithm Development and Validation","year":2021,"lang":"en","type":"article","venue":"JMIR Research Protocols","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University; Impact","funders":"","keywords":"Computer science; Machine learning; Hyperparameter; Artificial intelligence; Relevance (law); Identification (biology); Deep learning; Protocol (science); Data mining; Algorithm; Information retrieval; Medicine","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.06163774,0.00361576,0.001924919,0.004372644,0.002182787,0.003031754,0.004890768,0.003793097,0.03998923],"category_scores_gemma":[0.1318222,0.002461275,0.003745146,0.002699843,0.002933876,0.00237783,0.004156449,0.006215428,0.01188073],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003904037,"about_ca_system_score_gemma":0.02302173,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003292405,"about_ca_topic_score_gemma":0.005222504,"domain_scores_codex":[0.9797934,0.009934306,0.003650106,0.002474754,0.00357432,0.0005731297],"domain_scores_gemma":[0.9221389,0.03770213,0.004239684,0.01551418,0.01918918,0.001215961],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.01168103,0.004567561,0.007387927,0.02195258,0.001913119,0.001173937,0.001476377,0.06810937,0.02280859,0.0320312,0.1054081,0.7214901],"study_design_scores_gemma":[0.02962886,0.009092642,0.01045769,0.01432345,0.002415429,0.002552607,0.0009941993,0.3186037,0.09707118,0.1058252,0.4080134,0.001021589],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"protocol","genre_scores_codex":[0.009683521,0.001302835,0.6563973,0.001257747,0.000473937,0.3002316,0.01729178,0.007130933,0.006230304],"genre_scores_gemma":[0.009791954,0.0003494735,0.580285,0.0003977739,0.00004085338,0.4024033,0.004810309,0.0003715293,0.001549822],"genre_candidate":"protocol","genre_consensus":null,"teacher_disagreement_score":0.9383622,"threshold_uncertainty_score":0.3259754,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3789286136609896,"score_gpt":0.5897857567159527,"score_spread":0.2108571430549631,"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."}}