{"id":"W4311617908","doi":"10.1101/2022.11.30.22282946","title":"Discovering Social Determinants of Health from Case Reports using Natural Language Processing: Algorithmic Development and Validation","year":2022,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Topic Modeling","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Ontario; University of Toronto","funders":"Institute of Health Services and Policy Research; Canadian Institutes of Health Research","keywords":"Computer science; Benchmark (surveying); Natural language processing; Artificial intelligence; Social media; Social determinants of health; Annotation; Key (lock); Information extraction; Set (abstract data type); Information retrieval; Health care; Data science; Machine learning; World Wide Web; Political science","routes":{"ca_aff":true,"ca_fund":true,"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":[],"consensus_categories":[],"category_scores_codex":[0.0007040531,0.0001766562,0.0003517436,0.0001065222,0.0002932335,0.0001224885,0.0002384595,0.00006608454,0.000003447558],"category_scores_gemma":[0.0000311339,0.0001832748,0.00004095894,0.0001010115,0.00002172339,0.0001956904,0.001577989,0.0003141165,8.762862e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001960637,"about_ca_system_score_gemma":0.0005695535,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001312342,"about_ca_topic_score_gemma":0.00006237654,"domain_scores_codex":[0.9980573,0.0001216427,0.0006507575,0.000613635,0.0003483439,0.0002083178],"domain_scores_gemma":[0.998763,0.00002150169,0.0007848647,0.0003528727,0.00003531567,0.00004243025],"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":[0.000004282324,0.00006042398,0.06076115,0.0009342009,0.00004175082,0.007167581,0.08657976,0.001797395,0.0006927582,0.00001552699,0.000002178911,0.841943],"study_design_scores_gemma":[0.0002419929,0.00001457362,0.02132262,0.0004546835,0.00002335795,0.001996784,0.001368353,0.9683231,0.005246306,0.0003457396,0.00006489435,0.0005976615],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8489758,0.001308151,0.1488035,0.00003980573,0.0006291444,0.0001863264,0.000005080808,0.00004817136,0.000003984974],"genre_scores_gemma":[0.8839068,0.000002875041,0.1159414,0.00001790292,0.00007637931,0.00001497546,0.00001518787,0.00001274913,0.00001172059],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9665256,"threshold_uncertainty_score":0.7473729,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05517938981622243,"score_gpt":0.335244282657424,"score_spread":0.2800648928412016,"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."}}