{"id":"W3032372024","doi":"10.2196/19509","title":"Machine Learning to Detect Self-Reporting of Symptoms, Testing Access, and Recovery Associated With COVID-19 on Twitter: Retrospective Big Data Infoveillance Study","year":2020,"lang":"en","type":"article","venue":"JMIR Public Health and Surveillance","topic":"Mental Health via Writing","field":"Psychology","cited_by":155,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Pandemic; Coronavirus disease 2019 (COVID-19); Computer science; Topic model; Public health; Artificial intelligence; Social media; Public health surveillance; Medicine; Machine learning; Natural language processing; Information retrieval; World Wide Web; Disease; Infectious disease (medical specialty)","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":[0.001631152,0.0003843851,0.0004017203,0.001386855,0.0005331686,0.0009090218,0.000450987,0.0006902536,0.0009905164],"category_scores_gemma":[0.009830348,0.0002486254,0.0005069833,0.001183979,0.0003034903,0.001151141,0.00083262,0.0008673465,0.001047033],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000515996,"about_ca_system_score_gemma":0.000454541,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005892013,"about_ca_topic_score_gemma":0.007115093,"domain_scores_codex":[0.9987565,0.0004546322,0.0001422318,0.0002677855,0.0001936965,0.0001850223],"domain_scores_gemma":[0.9941314,0.002834621,0.001413213,0.0005411941,0.0006974022,0.0003821818],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0002630427,0.0002665041,0.9749426,0.0001211178,0.00007471095,0.0002255361,0.001432296,0.0006281583,0.001515938,0.000173331,0.003119556,0.01723718],"study_design_scores_gemma":[0.00001630006,0.0003398706,0.9454679,0.0001107265,0.0001062914,0.0006386464,0.005183788,0.0396304,0.0021614,0.0005831035,0.005708307,0.00005322919],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9888732,0.000166425,0.001740979,0.0005039699,0.00003935555,0.0001269192,0.007427838,0.00005898864,0.001062375],"genre_scores_gemma":[0.9888679,0.0001224603,0.002579978,0.0001797767,0.0000684855,0.0002007404,0.007312219,0.00001934336,0.0006491133],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005892013,"threshold_uncertainty_score":0.01171541,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1749465264901289,"score_gpt":0.4103833435911393,"score_spread":0.2354368171010104,"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."}}