{"id":"W4416431165","doi":"10.2196/73822","title":"Identifying Biomedical Entities for Datasets in Scientific Articles: 4-Step Cache-Augmented Generation Approach Using GPT-4o and PubTator 3.0","year":2025,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Metadata; Workflow; Data integration; Metadata modeling; Identification (biology)","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":[0.01153455,0.001366314,0.000933392,0.009876101,0.001163999,0.00399982,0.001682508,0.001354565,0.009453117],"category_scores_gemma":[0.043995,0.0008105112,0.003429459,0.005304206,0.0008326902,0.002724658,0.004804059,0.001348595,0.004508476],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001394183,"about_ca_system_score_gemma":0.005052039,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003888492,"about_ca_topic_score_gemma":0.01142248,"domain_scores_codex":[0.9949366,0.001909524,0.0007785523,0.001214158,0.001011902,0.000149236],"domain_scores_gemma":[0.9675792,0.02061834,0.002756755,0.004375264,0.004269532,0.0004008852],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00135281,0.0003930122,0.05550992,0.01152791,0.0013887,0.00244999,0.006169862,0.01658857,0.04263967,0.02483241,0.1199143,0.7172329],"study_design_scores_gemma":[0.0008065783,0.0006993207,0.06373636,0.002497867,0.003003983,0.004235582,0.003769148,0.3572977,0.07168408,0.08495892,0.4067307,0.0005797214],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04924658,0.00239788,0.821237,0.002226888,0.0005328559,0.003531361,0.06226589,0.04953568,0.00902592],"genre_scores_gemma":[0.07739358,0.0004864255,0.8489916,0.0003555154,0.00009968441,0.002406248,0.06627738,0.002039397,0.001950217],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01153455,"threshold_uncertainty_score":0.0610013,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1455260598801369,"score_gpt":0.4481222549381302,"score_spread":0.3025961950579932,"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."}}