Evaluation of Pronunciation by means of Automatic Speech Recognition System for Computer Aided Indonesian Language Learning

摘要

Utilizing of speech recognizers as pronunciation evaluator is discussed in this work. The main reason is to put application of speech technology in assisting Indonesian learning as a part of on-going open learning environment. With the advanced technology of speech recognizers, the possibilities of applications become wider. Dealing with isolated word recognizer and medium-sized database, evaluation of pronunciation of learners based on prosody information is the objective. Two kinds of word set from non-native speakers are used to evaluate automatic pronunciation recognizer: mispronunciation by mistake and those by linguistical inheritance. The quality of evaluation is confirmed by human judgment. In addition, what kinds of mispronunciation they would do in Indonesian is also investigated. Herein it explores the possibility of using a Hidden Markov Model HTK tools recognizer and of obtaining correct prosody information in order to implement it in a system to aid in Indonesian language learning as L2.