AI-Powered Adaptive Pronunciation Feedback Systems for Non-Native Students in Multilingual University Contexts

Abstract

New technologies of artificial intelligence evoke in mobile learners the need for reconstruction of the existing methods of language acquisition of the English language by native Bangla speakers. However, in multilingual contexts of the higher education sector, speaking and collaborating skills remain as obstacles. This specific research examines the impact of an adaptive artificial intelligence (AI) model integrated with real-time (what the author refers to as instantaneous) personalised speech analytics, instantaneous speech pronunciation assessment, and correction feedback (at the moment) to targeted users for the purpose of improving users' verbal communication skills. A mixed methods approach (qualitative and quantitative) provides the researcher with the means to analyse the data and accomplish the objectives of the research. It follows then, the authors present the findings of a five-week intervention research with students numbered up to one hundred and forty-two non-native speakers of the Bangla language. Participants of the research were studied for phonemic attainment as well as for degrees of foreign language (or spoken English) anxiety.