{"id":"W4385572774","doi":"10.18653/v1/2022.emnlp-main.89","title":"Geographic Citation Gaps in NLP Research","year":2022,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Citation; Publication; Field (mathematics); Computer science; Diversity (politics); China; Data science; Artificial intelligence; World Wide Web; Geography; Political science","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":[0.001550099,0.00003936402,0.00004645762,0.0006613971,0.0001794255,0.00009113536,0.0008737213,0.00001772295,0.00007064697],"category_scores_gemma":[0.00007271137,0.00003681743,0.00001593075,0.002107179,0.00002440512,0.0002753589,0.0006561574,0.0003851287,0.000007449923],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006469793,"about_ca_system_score_gemma":0.00004381198,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002187125,"about_ca_topic_score_gemma":0.00004445404,"domain_scores_codex":[0.9987913,0.0002046651,0.00009625079,0.0002220871,0.0004849069,0.00020074],"domain_scores_gemma":[0.9995295,0.00009839626,0.00001787305,0.0002642629,0.00006725085,0.00002276126],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000007227925,0.0001078868,0.006006091,0.00001763836,0.000002475599,0.00008783666,0.001692403,0.00003575258,0.004002805,0.9226335,0.007727135,0.05767922],"study_design_scores_gemma":[0.0002285595,0.0002123115,0.00399401,0.00001545818,5.363486e-7,0.00002654844,0.0003220941,0.02322209,0.007642291,0.9579012,0.006212322,0.000222618],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2530992,0.01519709,0.6358305,0.02808795,0.0008693488,0.001593764,0.000004502169,0.004885624,0.06043205],"genre_scores_gemma":[0.7982727,0.000004935922,0.2010667,0.0002137783,0.000008261643,0.00006334407,0.000001393439,0.000003028464,0.0003659271],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5451735,"threshold_uncertainty_score":0.1673214,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04003137957396832,"score_gpt":0.3559137284766063,"score_spread":0.315882348902638,"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."}}