{"id":"W3198457396","doi":"10.18653/v1/2021.emnlp-main.259","title":"Learning from Multiple Noisy Augmented Data Sets for Better Cross-Lingual Spoken Language Understanding","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Benchmark (surveying); Focus (optics); Noise (video); Code (set theory); Artificial intelligence; Training set; Machine learning; Spoken language; Resource (disambiguation); Noise reduction; Natural language processing; Speech recognition; Image (mathematics); Set (abstract data type)","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":["metaresearch","metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001854907,0.0004763986,0.0006907755,0.0002482856,0.0003758682,0.0009428226,0.003264668,0.0003163807,0.00003166546],"category_scores_gemma":[0.009259205,0.0003598461,0.0001797755,0.001448199,0.0002141622,0.00139279,0.002294722,0.001528368,0.000001662672],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002649171,"about_ca_system_score_gemma":0.0003164459,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006484587,"about_ca_topic_score_gemma":0.00003291322,"domain_scores_codex":[0.9960976,0.0002089891,0.0007934426,0.00145903,0.0007228011,0.0007181902],"domain_scores_gemma":[0.9965478,0.001376985,0.0006789412,0.0006639127,0.0006115467,0.0001208398],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002259188,0.0002145348,0.01199978,0.0006084258,0.00008400754,0.00005283427,0.01433812,0.00001253613,0.2832991,0.001548971,0.0001231851,0.6874926],"study_design_scores_gemma":[0.001700248,0.0001119776,0.001034267,0.002605308,0.0000780221,0.00003666307,0.007955085,0.43087,0.5311243,0.02348682,0.00008126267,0.0009160277],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4408099,0.02410701,0.5258783,0.005320973,0.0009976893,0.0012054,0.0001043915,0.0007135327,0.0008628137],"genre_scores_gemma":[0.5306469,0.00001000721,0.4686252,0.0003963271,0.0001051019,0.00002167561,0.00003827536,0.00002974012,0.0001267683],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6865766,"threshold_uncertainty_score":0.9998854,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1489856121254087,"score_gpt":0.4560963868956432,"score_spread":0.3071107747702345,"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."}}