General Design of Automatic Correction System for English Pronunciation Errors Based on DTW Algorithm

Abstract

In today’s world, economic cooperation is the general tendency, and almost people are attending the spoken English. Advances in learning have made it easier to spoken language, but the key to find and correct wrong words has shortcomings. The purpose of this paper is to design an automatic correction system for English pronunciation errors based on DTW algorithm. In the experiment, based on the principle of DTW algorithm, students were tested to read monosyllabic, disyllabic and multi-syllable words. The experimental results show that the mean of sample T(M=81, 0215) is basically equal to the mean of sample G(M=81, 03120). After analyzing the scoring, it shows that there is little difference between the two.