{"id":"W2989812388","doi":"10.26615/978-954-452-056-4_088","title":"Turning Silver into Gold: Error-Focused Corpus Reannotation with Active Learning","year":2019,"lang":"en","type":"article","venue":"","topic":"Machine Learning and Algorithms","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Computer Research Institute of Montréal","funders":"","keywords":"Computer science; Artificial intelligence; Selection (genetic algorithm); Natural language processing; Gold standard (test); Quality (philosophy); Factor (programming language); Random forest; Machine learning; Statistics","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":[0.0002272734,0.0001616258,0.000166744,0.0001197038,0.0001241753,0.0001731778,0.0003884596,0.00005351214,0.00009214848],"category_scores_gemma":[0.00005598351,0.0001211759,0.00004056126,0.0004117944,0.00002313177,0.0006187073,0.0001274518,0.0004456129,0.0004215763],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004754042,"about_ca_system_score_gemma":0.00005464531,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003371041,"about_ca_topic_score_gemma":0.00001897373,"domain_scores_codex":[0.9986984,0.000122676,0.0001436029,0.0004433862,0.0003297584,0.000262174],"domain_scores_gemma":[0.9992057,0.0001528661,0.0001396651,0.0003228112,0.00009843275,0.00008052982],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00008297654,0.0000700232,0.02867849,0.0000394581,0.00007498563,0.00003726291,0.01182675,0.07163978,0.006775368,0.01592015,0.0003109209,0.8645439],"study_design_scores_gemma":[0.001349396,0.0007143798,0.02472732,0.00007693862,0.000009889948,0.00003076598,0.0004557126,0.9577521,0.003122448,0.0006104472,0.01064748,0.0005031003],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.564136,0.00002254224,0.4007674,0.0007048806,0.0002988084,0.000195542,1.768182e-7,0.0006377929,0.03323683],"genre_scores_gemma":[0.9158559,0.000001469979,0.06850869,0.0001728207,0.00005638811,0.000006330319,0.000005519653,0.00001656349,0.01537627],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8861123,"threshold_uncertainty_score":0.5418651,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006310426710666149,"score_gpt":0.2315417910793546,"score_spread":0.2252313643686885,"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."}}