{"id":"W4386588264","doi":"10.1186/s44247-023-00035-y","title":"Discovering social determinants of health from case reports using natural language processing: algorithmic development and validation","year":2023,"lang":"en","type":"article","venue":"BMC Digital Health","topic":"Food Security and Health in Diverse Populations","field":"Health Professions","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; York University; Public Health Ontario; University of Toronto","funders":"Institute of Health Services and Policy Research; Canadian Institutes of Health Research","keywords":"Computer science; Natural language processing; Artificial intelligence; Benchmark (surveying); Social determinants of health; Social media; Annotation; Information extraction; Set (abstract data type); Key (lock); 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01595314,0.001147919,0.0006228758,0.005437044,0.0009289716,0.002020901,0.002574703,0.001726223,0.002334225],"category_scores_gemma":[0.04793646,0.0004989813,0.001403475,0.001978547,0.001497761,0.001773857,0.002262397,0.001576788,0.0008791315],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001209045,"about_ca_system_score_gemma":0.002750613,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005412149,"about_ca_topic_score_gemma":0.007232873,"domain_scores_codex":[0.9907573,0.005049626,0.001190364,0.001735601,0.001088738,0.0001783842],"domain_scores_gemma":[0.9159984,0.07443506,0.002030852,0.003386228,0.003839961,0.0003095498],"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":[0.0006173495,0.001572325,0.06219386,0.002344179,0.0005586396,0.002434123,0.002217476,0.1544697,0.008629546,0.00722492,0.01361982,0.7441181],"study_design_scores_gemma":[0.0001592836,0.0001645949,0.01140683,0.0002993931,0.0001410657,0.0008439361,0.0008351812,0.9613104,0.006378665,0.0126779,0.00574119,0.00004153387],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2139913,0.001599023,0.7657186,0.001608093,0.0001467919,0.003637851,0.005013099,0.005434379,0.002850735],"genre_scores_gemma":[0.267474,0.0003706215,0.720841,0.0002064051,0.00008076277,0.00162338,0.008823638,0.00009818217,0.0004820696],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01595314,"threshold_uncertainty_score":0.08436924,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2337726519686488,"score_gpt":0.4902910702304957,"score_spread":0.256518418261847,"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."}}