지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
이용수
초록· 키워드
오류제보하기
This paper presents a method of automatically assessing non-native speech produced by Korean learners of English. After an input utterance is processed by an automatic aligner to produce information on phoneme and word boundaries, the key modules are designed to output separate scores in each level of segmentals and prosody. For segmental evaluation, the HMM log likelihood scoring method has been adopted to properly compare input speech with native pronunciation. For prosodic scoring, rhythm metrics are employed to capture the rhythm characteristics of the learners' native language that are expected to remain in their English utterances. The results of correlation analyses between the automatic scores and human raters' scores suggest that the proposed method can be further applied to enhance performance of pronunciation self-education tools for EFL learners.
목차
등록된 정보가 없습니다.
참고문헌 (0)
참고문헌 신청이 논문의 저자 정보
최근 본 자료
댓글(0)
0