{"id":"W4252876123","doi":"10.2196/preprints.32355","title":"Physical Activity, Sedentary Behavior, and Sleep on Twitter: Multicountry and Fully Labeled Public Data Set for Digital Public Health Surveillance Research (Preprint)","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Data-Driven Disease Surveillance","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Social media; Computer science; Metadata; Data set; Public health; Public health surveillance; Set (abstract data type); Data quality; Data science; World Wide Web; Medicine; Engineering; Artificial intelligence","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.002471331,0.0005542672,0.0006540215,0.002158929,0.00107133,0.001021804,0.00115387,0.001534423,0.003018935],"category_scores_gemma":[0.008671807,0.0002745803,0.0009156566,0.002596058,0.0007443265,0.001593065,0.002363019,0.001446807,0.002475394],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001076601,"about_ca_system_score_gemma":0.001335671,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02362685,"about_ca_topic_score_gemma":0.03496445,"domain_scores_codex":[0.9980054,0.000767368,0.0002035946,0.0003863669,0.0003664958,0.0002707122],"domain_scores_gemma":[0.9912986,0.00312019,0.001194408,0.00225383,0.001445629,0.0006873432],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001397762,0.002009974,0.6042425,0.001581662,0.0005596283,0.001570601,0.002710911,0.01057564,0.006819818,0.003451364,0.2469776,0.1181026],"study_design_scores_gemma":[0.0001699806,0.0004929625,0.8405517,0.0003777927,0.0002267002,0.0007894033,0.00737384,0.05214421,0.004460242,0.002790252,0.09041163,0.0002112391],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"dataset","genre_scores_codex":[0.6478884,0.0006025535,0.006046261,0.004852482,0.0005440525,0.0004539031,0.3348718,0.0007192239,0.004021375],"genre_scores_gemma":[0.4507342,0.0003833036,0.01877984,0.0007287669,0.0005058797,0.001311992,0.5244637,0.0001125206,0.002979866],"genre_candidate":"dataset","genre_consensus":null,"teacher_disagreement_score":0.02362685,"threshold_uncertainty_score":0.04697865,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.257005036698103,"score_gpt":0.443589629380032,"score_spread":0.186584592681929,"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."}}