{"id":"W4399500684","doi":"10.2196/59794","title":"Ethics of the Use of Social Media as Training Data for AI Models Used for Digital Phenotyping","year":2024,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Digital Mental Health Interventions","field":"Psychology","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Preprint; Social media; Artificial intelligence; Psychology; Training set; Training (meteorology); Computer science; Data science; Machine learning; World Wide Web","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":["metaresearch","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.40726,0.0007587232,0.001063618,0.001295629,0.005568319,0.01006844,0.00548444,0.01837173,0.002782023],"category_scores_gemma":[0.5382128,0.0009981371,0.001328411,0.001054007,0.04468175,0.01039818,0.008322935,0.0224281,0.001350522],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004455644,"about_ca_system_score_gemma":0.01606465,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003156796,"about_ca_topic_score_gemma":0.003719851,"domain_scores_codex":[0.408133,0.5442352,0.01456907,0.006522804,0.02370889,0.002831066],"domain_scores_gemma":[0.2772875,0.6270577,0.0122907,0.04857896,0.03094802,0.003837142],"domain_codex":"methods","domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0003258162,0.0001183413,0.00172921,0.001055195,0.00008300264,0.0006080957,0.09741955,0.001440342,0.0009529109,0.7629707,0.06708664,0.06621026],"study_design_scores_gemma":[0.0001665375,0.0002966742,0.001026993,0.007410856,0.00009074984,0.0009827845,0.02170654,0.006570472,0.003841885,0.4311675,0.5265309,0.0002080047],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"commentary","genre_gemma":"empirical","genre_scores_codex":[0.01501022,0.002860409,0.1589975,0.766946,0.008532632,0.002091378,0.0003079878,0.0001601238,0.04509391],"genre_scores_gemma":[0.4934265,0.003973152,0.1485084,0.3113901,0.008637601,0.01124416,0.0002605309,0.0003045044,0.02225512],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9816283,"threshold_uncertainty_score":0.7309539,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6785152226372546,"score_gpt":0.6111623121444573,"score_spread":0.06735291049279735,"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."}}