Abstract:
This project aims to create a comprehensive, self-directed framework for interactive pronunciation learning in English and Mandarin, leveraging Corpus and AI technologies. Complementing this framework, the initiative also produces relevant teaching materials. The foundation of this endeavour is two speech corpora from PI’s previous TDG projects: The Spoken Corpus of Hong Kong Learners of Mandarin (T0150, 2015-2017) and The Spoken English Corpus of Chinese and Non-Chinese Learners in Hong Kong (T0200; 2018-2020). These resources provide authentic speech examples from both native and non-native Hong Kong speakers, helping learners identify key phonological characteristics of both languages.
To enhance pronunciation instruction, conversational and automatic speech recognition (ASR) AI tools, including Copilot, Murf, and Immersive Reader, are integrated into this corpus-based learning environment. This integration positions the project at the forefront of AI-assisted pronunciation learning and equips educators with practical, innovative teaching tools.
The initiative targets multilingual Chinese learners, addressing pronunciation challenges in their second (L2) and third (L3) languages, thus significantly improving their communication skills. Additionally, the materials developed will support teacher training, fostering collaboration between English and Mandarin educators to collectively address pronunciation issues common among Cantonese-speaking learners.
Code:
T0274
Principal Project Supervisors:
Keywords provided by authors:
- English pronunciation learning
- Mandarin pronunciation learning
- Pronunciation teaching and learning in the multilingual context
- Self-directed pronunciation learning
- AI and corpus-aided pronunciation teaching
Start Date:
01 Nov 2023
End Date:
31 Dec 2024
Status:
Completed
Result:
This project developed a comprehensive framework, teaching kits, and a training program centred around AI and corpus-aided pronunciation learning for both English and Mandarin. Evaluations conducted with postgraduate and undergraduate students, as well as in-service teachers, yielded positive results. Participants reported a significant improvement in their understanding and application of corpus and AI tools, leading to increased confidence in their pronunciation abilities and a heightened awareness of accented English and Mandarin. Post-training assessments demonstrated substantial gains in English segmental, suprasegmental, and prosodic features, coupled with improved accuracy in identifying Mandarin tones and applying pronunciation rules, though AI tool adoption for Mandarin remained lower than for English. A dedicated public website was launched to disseminate project resources, including the online training program, thereby expanding access to these innovative tools. Feedback from participants underscored the value of authentic examples derived from corpora and the personalized feedback provided by AI tools, particularly in the context of accented speech. The project successfully integrated cutting-edge technology into pronunciation instruction, offering valuable resources and demonstrating the potential for enhanced learning outcomes.
Impact:
This project successfully integrated AI and corpus-based tools into the landscape of pronunciation teaching and learning, showcasing the transformative potential of these technologies for enhancing pronunciation skills and cultivating positive attitudes towards accent diversity. The development of a comprehensive framework, alongside accompanying teaching kits and a structured training program, provides invaluable resources for both educators and learners. The positive evaluation results, reflecting increased learner confidence and demonstrably improved pronunciation performance in both English and Mandarin, underscore the efficacy of this approach. The deliberate focus on accented English and Mandarin directly contributes to fostering intercultural communication competence and addresses the inherent challenges posed by pronunciation variation in an increasingly globalized world. The establishment of a public website significantly broadens the project’s reach and facilitates wider adoption of these innovative methodologies. The project’s significant contribution lies in demonstrating the potential of technology to personalize pronunciation learning, accommodate diverse accents, and provide readily accessible and valuable resources for both learners and educators.
Deliverables:
Seminars/ Presentations/ Sharing Sessions
Chen, H. C. (2023, 14 December). 如何將自主研發的口語語料庫應用於語音教學和學習 [Online seminar presentation]. Northwest Minzu University, China. (Number of participants: 40 students and staff members)
Chen, H. C., Han, Q., Chan, C. H. J., Tian, J., & Zhou, X. (2024, 20 September). AI and corpus-aided self-directed interactive English and Mandarin pronunciation [Online workshop presentation]. Education University of Hong Kong.
Chen, H. C., & Chan, C. H. J. (2024, 11 July). Enhancing intercultural communication pedagogy: Raising awareness of accented English with corpus and AI technology among pre-service teachers in Hong Kong [Conference presentation]. International Postgraduate Roundtable and Research Forum cum Summer School 2024, Education University of Hong Kong.
Chan, C. H. J., & Chen, H. C. (2024). Enhancing English pronunciation production and perception through AI and corpus-aided pedagogy [Conference presentation submitted]. International Conference on Learning and Teaching 2025, Education University of Hong Kong.
Teaching and Learning Resources/ Materials (including online resources)
A project website with AI and speech corpus-aided self-directed interactive teaching and learning framework and teaching resources (https://corpus.eduhk.hk/L3PLT/index.php/aicorpus-integration/)
Resources for teaching English and Mandarin pronunciation on the Canvas learning management system
Financial Year:
2023-24
Type:
TDG