{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01349263,0.000108686,0.0001856175,0.00005907966,0.0004806491,0.0002023231,0.0004622277,0.0002651303,0.0000064296],"category_scores_gemma":[0.006335419,0.00006219671,0.00005837377,0.0006136335,0.0007420295,0.000006683354,0.0005696096,0.000636454,0.000004553278],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002265139,"about_ca_system_score_gemma":0.0003252939,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001671329,"about_ca_topic_score_gemma":0.000009334253,"domain_scores_codex":[0.9946141,0.002807059,0.0007158552,0.0005883247,0.0008714357,0.0004032519],"domain_scores_gemma":[0.9971819,0.0009845907,0.0001269454,0.0005539977,0.001013894,0.0001386588],"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":[0.001052639,0.001185441,0.001290587,0.0003895503,0.00008785633,0.000002546834,0.00167177,0.000004311632,0.09935271,0.0005741235,0.009767361,0.8846211],"study_design_scores_gemma":[0.001944019,0.00100719,0.01812008,0.0003893846,0.000002283882,0.000004223885,0.0009890364,0.0004186918,0.2199498,0.001390917,0.7556463,0.0001380126],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"protocol","genre_gemma":"protocol","genre_scores_codex":[0.02098232,0.00003187832,0.04064234,0.003459286,0.0000123141,0.9347681,0.00003471563,0.0000209804,0.000048035],"genre_scores_gemma":[0.005458591,0.000001478809,0.02869756,0.00004829776,0.0003018268,0.9649137,0.0002346329,0.00001390949,0.000330039],"genre_candidate":"protocol","genre_consensus":"protocol","teacher_disagreement_score":0.8844831,"threshold_uncertainty_score":0.7584546,"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."}}