Modelling of the perception of English sentence stress for computer-assisted language learning

For learning foreign language pronunciation, prosodic features are important as much as, or more than segmental features. For Japanese speakers, one of difficulties to learn English pronunciation is rhythm because of the differences between two languages: mora-timing rhythm and stress-timing rhythm. In order to correct errors in rhythm, the method of evaluating sentence stress that constitutes rhythm is very significant. In this paper, we present a method of automatic detecting sentence stress syllables for the evaluation criterion. Using a linear discriminant function of pitch, intensity and vowel duration, about 90% of the syllables were correctly detected as to sentence stress. Also we analyzed the different and common characteristics among different English native speakers. The results revealed that the perception of the sentence stress among 11 native speakers had general agreement with respect to how to integrate three features.