{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01571393,0.002313674,0.002083522,0.005926617,0.002808356,0.005166221,0.005108056,0.004314053,0.00408636],"category_scores_gemma":[0.04845278,0.0008967368,0.001050091,0.004121357,0.003486675,0.006176082,0.008790381,0.004117757,0.005326563],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001293705,"about_ca_system_score_gemma":0.002395316,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0037414,"about_ca_topic_score_gemma":0.008819125,"domain_scores_codex":[0.9847355,0.006105256,0.001553531,0.003277254,0.003602094,0.0007264307],"domain_scores_gemma":[0.949343,0.02460754,0.002485663,0.01386996,0.008825806,0.0008680278],"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":[0.001311408,0.0008874173,0.0085799,0.001927131,0.0003087824,0.0009596467,0.005325489,0.01994023,0.04009031,0.01826953,0.09176975,0.8106304],"study_design_scores_gemma":[0.0004623293,0.0008228208,0.008431097,0.000793231,0.0003039789,0.001399545,0.00516728,0.4739498,0.210345,0.09502251,0.2028664,0.0004360173],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1324751,0.00336764,0.7948116,0.003315496,0.002328927,0.001212241,0.008139175,0.03837734,0.01597241],"genre_scores_gemma":[0.2520581,0.000566268,0.7043306,0.001222236,0.0003308359,0.001280645,0.02178714,0.004883224,0.01354096],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01571393,"threshold_uncertainty_score":0.08310419,"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."}}