{"id":"W4383724340","doi":"10.1080/19331681.2023.2231436","title":"Categorizing political campaign messages on social media using supervised machine learning","year":2023,"lang":"en","type":"article","venue":"Journal of Information Technology & Politics","topic":"Social Media and Politics","field":"Social Sciences","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Universidade Federal de Minas Gerais; Columbia University; University of Liverpool; University of Glasgow; University of Guelph; John S. and James L. Knight Foundation","keywords":"Politics; Artificial intelligence; Computer science; Social media; Journalism; The Internet; Presidential system; Public relations; Sociology; Media studies; World Wide Web; Political science; Law","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007795076,0.0001474708,0.0003297287,0.001228246,0.000916058,0.00009204258,0.0003652118,0.0004755457,0.00004366887],"category_scores_gemma":[0.004941316,0.0001435665,0.0001361145,0.0008404645,0.0005798041,0.0005973445,0.00005694341,0.0008475267,0.00008437701],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005211373,"about_ca_system_score_gemma":0.000701657,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002071129,"about_ca_topic_score_gemma":0.00002030826,"domain_scores_codex":[0.9974849,0.0001852966,0.0007233602,0.00006490966,0.0006948939,0.0008466502],"domain_scores_gemma":[0.9980137,0.000690903,0.0004430958,0.00008769299,0.0005226912,0.0002419335],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"qualitative","study_design_scores_codex":[0.000009902784,0.00002394301,0.008166313,0.00002305703,0.00004416108,0.00001941136,0.03681694,0.00004468785,0.0001003721,0.9508691,0.000733072,0.003149074],"study_design_scores_gemma":[0.002052076,0.0003250671,0.002703607,0.000163717,0.0002180374,0.0000912999,0.6629683,0.00156782,0.005908201,0.1846868,0.138598,0.0007171497],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9815274,0.00004366329,0.0004483668,0.01160486,0.00146763,0.0001182946,0.00002596506,0.0003247609,0.004438987],"genre_scores_gemma":[0.9975633,0.00008835102,0.0005208259,0.00058036,0.001169525,0.000002177498,0.00001444451,0.00001444365,0.00004659675],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7661823,"threshold_uncertainty_score":0.704567,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04813084366387452,"score_gpt":0.3408060996038219,"score_spread":0.2926752559399474,"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."}}