{"id":"W2971788743","doi":"10.1177/0165551519871828","title":"DelibAnalysis: Understanding the quality of online political discourse with machine learning","year":2019,"lang":"en","type":"article","venue":"Journal of Information Science","topic":"Social Media and Politics","field":"Social Sciences","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Quality (philosophy); Computer science; Politics; Artificial intelligence; Classifier (UML); Discourse analysis; Data science; Natural language processing; Epistemology; Linguistics; Political science","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003043029,0.00004347765,0.0001431348,0.0001765952,0.0003452403,0.00009780995,0.0003542668,0.00002586824,0.00004225108],"category_scores_gemma":[0.001732835,0.00002436494,0.0000576603,0.0007891743,0.0009560084,0.001946537,0.00002209736,0.0002058333,0.000005130171],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002184248,"about_ca_system_score_gemma":0.0009889876,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004541137,"about_ca_topic_score_gemma":0.000121348,"domain_scores_codex":[0.9978368,0.0001358768,0.0004635546,0.00003163924,0.001305068,0.0002270934],"domain_scores_gemma":[0.9981833,0.0004104448,0.0006588008,0.00007440116,0.0005368678,0.0001362262],"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":[0.00003533032,0.00002751752,0.1550808,0.00001403076,0.00002286047,3.228939e-7,0.07361791,0.0007354495,0.0001325151,0.769448,0.00001542562,0.0008698427],"study_design_scores_gemma":[0.0005738126,0.0002113816,0.02896717,0.00008436792,0.00005110873,0.000008213415,0.9597076,0.001776771,0.0002469556,0.005757222,0.002489544,0.0001258892],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.979515,0.0000140182,0.007388388,0.004690566,0.0002279653,0.00006554776,0.000003947659,0.000005062833,0.008089472],"genre_scores_gemma":[0.9992361,0.00001656256,0.0003699933,0.0002589897,0.00008508088,1.698012e-7,6.451513e-7,0.000001185986,0.00003124481],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8860896,"threshold_uncertainty_score":0.3522453,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07076107172106615,"score_gpt":0.4155091402430217,"score_spread":0.3447480685219556,"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."}}