https://scholars.lib.ntu.edu.tw/handle/123456789/498586
標題: | Toward unsupervised discovery of pronunciation error patterns using universal phoneme posteriorgram for computer-assisted language learning | 作者: | Wang, Y.-B. LIN-SHAN LEE |
關鍵字: | GMM-MDL; HAC; K-means; Pronunciation Error Pattern Discovery; Rand Index; Universal Phoneme Posteriorgram | 公開日期: | 2013 | 起(迄)頁: | 8232-8236 | 來源出版物: | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings | 會議論文: | 2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 | 摘要: | In Computer-Aided Pronunciation Training, we hope to specify the type of mispronunciation, or Error Pattern (EP), the language learner has made as a more effective feedback. But derivation of EPs usually requires expert knowledge and pedagogical experiences, which is not easy to obtain for each pair of target and native languages. In this paper we propose a preliminary framework toward unsupervised discovery of EPs from a corpus of learners' recordings. We use Universal Phoneme Posteriorgram, derived from Multi-Layer Perceptron trained with a corpus of mixed languages, as features to bring supervised knowledge into the unsupervised task. We also use Hierarchical Agglomerative Clustering algorithm to explore sub-segmental variation of phoneme segments for distinguishing EPs. We tested K-means (assuming known number of EPs) and Gaussian Mixture Model with minimum description length principle (estimating unknown number of EPs) for EP discovery. Preliminary experimental results illustrated the effectiveness of the proposed framework, although there is still a long way to go compared to human annotators. © 2013 IEEE. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/498586 https://www.scopus.com/inward/record.uri?eid=2-s2.0-84890507052&doi=10.1109%2fICASSP.2013.6639270&partnerID=40&md5=eab572e36bda2aae49e06b6b430a8f3f |
ISSN: | 15206149 | DOI: | 10.1109/ICASSP.2013.6639270 | SDG/關鍵字: | Error patterns; GMM-MDL; HAC; K-means; Posteriorgram; Rand index; Electrical engineering; Signal processing |
顯示於: | 電機工程學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。