{"id":"W1970536639","doi":"10.3366/cor.2007.2.1.97","title":"The Wenzhou Spoken Corpus","year":2007,"lang":"en","type":"article","venue":"Corpora","topic":"China's Ethnic Minorities and Relations","field":"Social Sciences","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"XPath; Markup language; XML; Computer science; Transcription (linguistics); Natural language processing; Information retrieval; Linguistics; Artificial intelligence; World Wide Web; XML database","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":[0.001842263,0.0003843027,0.0004482683,0.003069462,0.002022483,0.001424176,0.0007583881,0.0004001487,0.03135153],"category_scores_gemma":[0.004508547,0.0002614993,0.0001589749,0.006211312,0.000864343,0.001038273,0.002029757,0.000473971,0.005734429],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001676874,"about_ca_system_score_gemma":0.004514636,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.03013672,"about_ca_topic_score_gemma":0.03655652,"domain_scores_codex":[0.9988416,0.0003304677,0.0002402693,0.0001954991,0.0002917983,0.0001003833],"domain_scores_gemma":[0.9975219,0.0008438833,0.0001145188,0.0005917901,0.0007633068,0.000164598],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.0006259677,0.0001674419,0.02343214,0.003815094,0.0001277154,0.004588209,0.04336408,0.002000437,0.02417262,0.1066647,0.4101672,0.3808744],"study_design_scores_gemma":[0.00007324634,0.0000525585,0.04765501,0.0001812425,0.00004625161,0.0005949491,0.007318745,0.001332548,0.005476285,0.005507349,0.9316925,0.00006937113],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.37717,0.003249293,0.0323571,0.00338961,0.0008495188,0.003062657,0.4132438,0.002386761,0.1642913],"genre_scores_gemma":[0.5359937,0.001734209,0.03818204,0.0005852701,0.0002217128,0.005392179,0.3546733,0.0008304566,0.06238724],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.03135153,"threshold_uncertainty_score":0.1048813,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02606158654718032,"score_gpt":0.3246620743798674,"score_spread":0.2986004878326871,"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."}}