{"id":"W2534147738","doi":"10.1145/2983323.2983694","title":"Optimizing Nugget Annotations with Active Learning","year":2016,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada; University of Waterloo; Google","keywords":"Automatic summarization; Computer science; Annotation; Sentence; Process (computing); Sequence (biology); Precision and recall; Recall; Natural language processing; Artificial intelligence; Information retrieval; Programming language","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":[0.00005850138,0.00005199286,0.00005003047,0.0000410052,0.00008720373,0.00004713247,0.0002243603,0.00001564394,0.00003413306],"category_scores_gemma":[0.00002179358,0.00002907048,0.00001278091,0.0001068683,0.00001391993,0.0005301913,0.00007156329,0.00005036931,0.00004723798],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002254362,"about_ca_system_score_gemma":0.0000326178,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001122479,"about_ca_topic_score_gemma":0.000007272881,"domain_scores_codex":[0.9994782,0.00002079307,0.00006342297,0.0001938421,0.0001074322,0.0001363622],"domain_scores_gemma":[0.9996104,0.00007838629,0.00002896294,0.0001935008,0.00004958553,0.00003918072],"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":[0.000008932695,0.00002886047,0.001159132,0.000003725429,0.00003026923,0.00001541021,0.004533988,0.01706408,0.006107353,0.2108964,0.0001758459,0.759976],"study_design_scores_gemma":[0.002448829,0.0004169402,0.004345497,0.0002110824,0.00001477369,0.00007025969,0.001097924,0.9271314,0.0373075,0.008425091,0.01755997,0.0009707731],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01663316,0.000004585958,0.9644007,0.001369128,0.00003833924,0.00003939058,1.430834e-7,0.000197701,0.01731682],"genre_scores_gemma":[0.6492209,0.000001475495,0.3486696,0.000085337,0.00001769864,0.000004944297,1.030355e-7,0.000003076584,0.00199681],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9100673,"threshold_uncertainty_score":0.118546,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01746247009185028,"score_gpt":0.236603102277309,"score_spread":0.2191406321854587,"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."}}