{"id":"W3172401221","doi":"10.1017/pan.2021.15","title":"Multi-Label Prediction for Political Text-as-Data","year":2021,"lang":"en","type":"article","venue":"Political Analysis","topic":"Topic Modeling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Centre for Social Innovation","funders":"Compute Canada; National Science Foundation","keywords":"Computer science; Machine learning; Artificial intelligence; Code (set theory); Set (abstract data type); Supervised learning; Training set; Source code; Government (linguistics); Data set; Association (psychology); Natural language processing; Artificial neural network; Psychology","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002799367,0.0001359519,0.0003017361,0.0001595515,0.0001278784,0.000179384,0.000931517,0.0001021889,0.00006906146],"category_scores_gemma":[0.0009126934,0.0001290611,0.0001887801,0.0007677886,0.0000601903,0.0003294794,0.0006787383,0.0001348089,0.00006667244],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000976182,"about_ca_system_score_gemma":0.0001820024,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003219547,"about_ca_topic_score_gemma":0.00006256279,"domain_scores_codex":[0.9975669,0.00008371197,0.0003643323,0.0008165215,0.0003354473,0.0008330255],"domain_scores_gemma":[0.9973664,0.0003258665,0.00003100514,0.001550405,0.0002371888,0.0004891481],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000001454206,0.000158863,0.001703012,0.00001326341,0.0002968613,0.00001247851,0.00002235214,0.00003512662,0.0001273066,0.994516,0.0001030461,0.003010304],"study_design_scores_gemma":[0.0003307232,0.00001932613,0.001420077,0.000004752179,0.0004930602,0.00001164506,0.00004813323,0.962186,0.0004934835,0.03397176,0.0008885054,0.0001324895],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003248843,0.00006252677,0.9822071,0.01238574,0.0001377172,0.00007582192,0.0001426305,0.0001400489,0.001599593],"genre_scores_gemma":[0.6939118,0.000001641234,0.302993,0.001937613,0.0002345456,0.00001575509,0.0001302134,0.000007780583,0.0007676441],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9621509,"threshold_uncertainty_score":0.5262957,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0953883885612755,"score_gpt":0.3506028114180816,"score_spread":0.2552144228568061,"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."}}