{"id":"W4386035682","doi":"10.2196/49721","title":"Automatic Extraction of Research Themes in Epidemiological Criminology From PubMed Abstracts From 1946 to 2020: Text Mining Study","year":2023,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Criminal Justice and Corrections Analysis","field":"Social Sciences","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Prison; Epidemiology; Mental health; Public health; Ranking (information retrieval); Qualitative research; Criminal justice; Psychology; Applied psychology; Medicine; Criminology; Sociology; Social science; Psychiatry; Computer science; Pathology; Information retrieval","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":["metaresearch","bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.01940781,0.001783175,0.002778695,0.08966382,0.001911178,0.004698015,0.001990192,0.001965949,0.004316218],"category_scores_gemma":[0.09094965,0.0009492343,0.004522982,0.04542899,0.0009365679,0.004031383,0.003679835,0.001115589,0.00250331],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002699816,"about_ca_system_score_gemma":0.01221894,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004528279,"about_ca_topic_score_gemma":0.009552065,"domain_scores_codex":[0.9823704,0.00317701,0.009406353,0.002361136,0.002164201,0.000520987],"domain_scores_gemma":[0.8860322,0.07797454,0.0177144,0.00266975,0.0141336,0.001475621],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"observational","study_design_scores_codex":[0.002771856,0.0005155371,0.1283257,0.4128826,0.004715338,0.008509577,0.01016776,0.00181454,0.01852134,0.002771211,0.05605509,0.3529494],"study_design_scores_gemma":[0.001353501,0.001783767,0.3523549,0.1500556,0.03022453,0.01089253,0.02354966,0.01071368,0.01799987,0.009470888,0.3909094,0.0006916483],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.2129293,0.1328549,0.01863921,0.005047017,0.001083295,0.01258166,0.6051157,0.002188408,0.009560597],"genre_scores_gemma":[0.3908451,0.06834149,0.1683621,0.002115268,0.0008366529,0.0239774,0.3419333,0.0005042154,0.003084511],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9805922,"threshold_uncertainty_score":0.1026396,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3568395366399731,"score_gpt":0.5411001991556078,"score_spread":0.1842606625156347,"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."}}