{"id":"W4242624664","doi":"10.31219/osf.io/ndp8m","title":"Crowdsourcing for machine learning in public health surveillance: lessons learned from Amazon Mechanical Turk","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Mobile Crowdsensing and Crowdsourcing","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Agency of Canada; University of Calgary","funders":"Natural Sciences and Engineering Research Council of Canada; Cumming School of Medicine, University of Calgary; Public Health Agency; Public Health Agency of Canada","keywords":"Crowdsourcing; Computer science; Artificial intelligence; Machine learning; Convolutional neural network; Data science; Inference; Quality (philosophy); Citizen science; Context (archaeology); Deep learning; Leverage (statistics); Ground truth; Natural language processing; World Wide Web","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":[0.02163656,0.001000378,0.001302001,0.001401319,0.001868217,0.003751118,0.002676137,0.002884826,0.003879763],"category_scores_gemma":[0.08531834,0.000654062,0.0009499268,0.002263797,0.004352895,0.004672089,0.003628056,0.002523015,0.001372486],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003406581,"about_ca_system_score_gemma":0.003457662,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04467585,"about_ca_topic_score_gemma":0.03388768,"domain_scores_codex":[0.9855779,0.01062386,0.0003329534,0.001340784,0.001748228,0.0003763186],"domain_scores_gemma":[0.8988373,0.08639331,0.00173781,0.00655202,0.005401045,0.001078389],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001246701,0.0007334813,0.05256708,0.002353801,0.000460847,0.0009351373,0.007908128,0.1503515,0.002172562,0.1069126,0.07438464,0.5999735],"study_design_scores_gemma":[0.0001694503,0.0002650508,0.01959097,0.001029679,0.00008289589,0.0001997234,0.003126175,0.6084741,0.002156268,0.303109,0.06158595,0.0002107585],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2263063,0.04292689,0.5236466,0.1256444,0.002236229,0.001746313,0.003734208,0.0022014,0.07155775],"genre_scores_gemma":[0.8596829,0.005196939,0.1225303,0.005037132,0.001210327,0.000831591,0.001139519,0.0003020637,0.004069293],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.04467585,"threshold_uncertainty_score":0.1144264,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1054447398296236,"score_gpt":0.3199346824422148,"score_spread":0.2144899426125912,"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."}}