{"id":"W2775747321","doi":"","title":"WiNER: A Wikipedia Annotated Corpus for Named Entity Recognition","year":2017,"lang":"en","type":"article","venue":"International Joint Conference on Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":38,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Computer science; Natural language processing; Information retrieval; Artificial intelligence; Quality (philosophy); Named-entity recognition; Simple (philosophy); Training set; Range (aeronautics); Task (project management)","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":[0.001937892,0.001156001,0.0007606663,0.008492736,0.001331907,0.001221952,0.001727385,0.001216258,0.008344822],"category_scores_gemma":[0.01009449,0.0006307687,0.00062842,0.005291368,0.0005276428,0.003341444,0.001665736,0.001686168,0.006529647],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005825823,"about_ca_system_score_gemma":0.002104229,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008997921,"about_ca_topic_score_gemma":0.01863313,"domain_scores_codex":[0.9978318,0.0005427693,0.0003503999,0.0006255205,0.0005258772,0.0001236432],"domain_scores_gemma":[0.9922423,0.002378907,0.0006464412,0.001584079,0.002659208,0.0004891221],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0005942069,0.000692555,0.01005735,0.0038305,0.0002397882,0.001839053,0.001586921,0.006727236,0.03953732,0.01369613,0.6251814,0.2960176],"study_design_scores_gemma":[0.0001772107,0.0002330677,0.02000463,0.000480075,0.0001600957,0.001765128,0.0007425852,0.03939936,0.03722933,0.008827311,0.890723,0.0002581321],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.07785952,0.003006589,0.2524015,0.001451778,0.002208153,0.001692725,0.5728976,0.04618397,0.04229807],"genre_scores_gemma":[0.06207394,0.0006622871,0.2182993,0.0003362275,0.0002311401,0.001281319,0.7059507,0.00229339,0.008871692],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008997921,"threshold_uncertainty_score":0.02791625,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06018724801345868,"score_gpt":0.3217091971620433,"score_spread":0.2615219491485846,"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."}}