{"id":"W7132590463","doi":"","title":"Supporting SENĆOŦEN language documentation efforts with Automatic Speech Recognition","year":2025,"lang":"en","type":"article","venue":"NPARC","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Pipeline (software); Documentation; Vocabulary; Word error rate; Speech technology; Language model; Set (abstract data type); Variation (astronomy)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001770919,0.001558956,0.0009802454,0.002059659,0.0009632243,0.002039297,0.001339789,0.001071289,0.008242586],"category_scores_gemma":[0.006247476,0.0005735395,0.0008253271,0.001269628,0.0006259591,0.002175339,0.002198796,0.002439988,0.02094421],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009168747,"about_ca_system_score_gemma":0.003161526,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01170897,"about_ca_topic_score_gemma":0.01653471,"domain_scores_codex":[0.9979312,0.0004611711,0.0001637797,0.0005970747,0.000687143,0.0001595809],"domain_scores_gemma":[0.995455,0.001073231,0.0002429024,0.001073278,0.001973215,0.0001823244],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002609637,0.0001452136,0.002040474,0.0004991424,0.00005869554,0.0004372803,0.0006033466,0.007140971,0.05758305,0.002027391,0.04150509,0.8876983],"study_design_scores_gemma":[0.0001760991,0.000452321,0.007634828,0.0002692902,0.0001259168,0.001039562,0.001579524,0.5253731,0.2751084,0.009794322,0.178205,0.0002416224],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09798593,0.001437098,0.6670179,0.001390719,0.0005941885,0.000460093,0.007748341,0.1970793,0.02628649],"genre_scores_gemma":[0.3622409,0.001069899,0.5709748,0.0005432719,0.0002333519,0.0003824136,0.03783493,0.003782411,0.02293799],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01170897,"threshold_uncertainty_score":0.02757418,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01088251064359417,"score_gpt":0.2816661806138079,"score_spread":0.2707836699702137,"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."}}