{"id":"W3137010024","doi":"10.1162/tacl_a_00447","title":"Quality at a Glance: An Audit of Web-Crawled Multilingual Datasets","year":2022,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":167,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; Google (Canada)","funders":"Agence Nationale de la Recherche","keywords":"Computer science; USable; Audit; Natural language processing; Quality (philosophy); Artificial intelligence; Information retrieval; World Wide Web; Accounting","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":[0.0008875114,0.00009488399,0.0001897328,0.00008926034,0.0005031255,0.00002200382,0.0008587202,0.0000477147,0.00001285938],"category_scores_gemma":[0.001566438,0.00009121074,0.0001359563,0.0003749802,0.00003577304,0.00007443652,0.00009772754,0.0001745594,5.19751e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003728558,"about_ca_system_score_gemma":0.0002542236,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004269242,"about_ca_topic_score_gemma":0.00003349575,"domain_scores_codex":[0.9983432,0.0001840854,0.0004607408,0.0002138657,0.0006618442,0.0001362973],"domain_scores_gemma":[0.9971779,0.0007856539,0.0007580015,0.0003375986,0.0009098315,0.00003099284],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003399374,0.001879743,0.005995675,0.0003931012,0.0004356662,0.000001635489,0.003760186,0.7656395,0.004124791,0.2019991,0.005225493,0.01020511],"study_design_scores_gemma":[0.002193061,0.0003576371,0.003630752,0.00003403901,0.0001259193,0.000006118806,0.0001075539,0.8041309,0.01236331,0.1620325,0.01454772,0.0004705009],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01121246,0.0001524962,0.9812422,0.0003335915,0.001485828,0.0004672054,0.004789886,0.0002091201,0.0001072298],"genre_scores_gemma":[0.7642645,7.697161e-7,0.2352252,0.00005920379,0.00005585414,0.00002500295,0.0002065392,0.000007330138,0.0001555747],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7530521,"threshold_uncertainty_score":0.3869685,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02197055172116473,"score_gpt":0.3306852318074935,"score_spread":0.3087146800863287,"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."}}