A Voicebot with Pronunciation Feedback Mechanism for Rural Indian English Learners

In this current digital world, the proficiency in the English language can bring in more employment and economic opportunities. Vocabulary retention capacity is the major factor in developing the ability to speak and write for an individual. In addition, fluency in any language is also influenced by the pronunciation of words while speaking. The rural Indian population struggles to speak good English, and the lack of vocabulary prevents their opportunity in this digital world. In this work, we have introduced a digital voice bot that helps users learn English words and sentences with a special focus on improving pronunciation. The paper focuses on how NLP techniques, along with Whisper an Automatic Speech Recognition model (ASR), can be utilized to help students provide feedback on their learning and pronunciation. The work also discusses how the software is designed to introduce new vocabulary to learners in a hierarchical manner. The multi-level hierarchy presented in this work ensures that the user will be able to retain good vocabulary along with right pronunciation once he/she completes the different levels of bot software.