A corpus-based online pronunciation learning system: The pedagogical applications of spoken corpus in improving Hong Kong / Mainland university students' English pronunciation

This project aims to develop a corpus-based, online pronunciation learning system to help HKIEd’s teachers, learners and researchers have a better understanding of the major problems in learning pronunciation by Hong Kong and Mainland English learners. A learners’ pronunciation corpus will be established and analysed in order to (a) identify learners’ recurrent difficulties in using English segmental features and suprasegmental features accurately and appropriately and (b) suggest possible remedies that may reduce or eliminate such difficulties. The research findings and products will inform teaching and enhance teaching quality. The two major products of this project are: 1. A phonological corpus (10-12 hours recording data containing roughly 100,000 English words) of spoken English from Hong Kong and Mainland students at HKIEd, which contains data in the form of audio files as well as phonological annotations, with a special focus on three areas of segmental features (vowels, consonants and consonant clusters) and four areas of suprasegmental features (stress assignment on both word level and the utterance level, pause, intonation and speech rhythm); 2. An online learning system that will enable learners to study segmental and suprasegmental features through perception and production exercises designed by the team based on the corpus data collected in this study.