Automatic Speech Recognition and Pronunciation Training

Automatic speech recognition (ASR) is providing more speaking opportunities for second language learners. With its progress in more accurate error detecting, the application of ASR in language teaching holds huge potential to practice and improve pronunciation. The purpose of the study is to present the features and application of ASR in second language learning. This paper firstly focuses on presenting the available literature on pronunciation training. Then the effectiveness of the ASR in evaluating the pronunciation data and providing feedback is discussed by referring to some evidence. It is found that state-of-the-art ASR is conducive as a pronunciation measurement, but the feedback provided by ASR is less effective.