Conversational AI for Enhancing English Speaking Proficiency: A Mobile App

Authors

  • Areeg Fahad Rasheed Network Dept, Al-Nahrain University, Baghdad, Iraq
  • M. Zarkoosh Independent Researcher, Baghdad, Iraq

DOI:

https://doi.org/10.25124/ijait.v8i1.6960

Keywords:

Conversational AI, Chat GPT, DialoGPT Blenderbot, machine learning

Abstract

As one of the most widely spoken languages globally, English plays a pivotal role in academia, digital communication, and numerous professional domains. English proficiency is vital for academic pursuits, international communication, and careers. The profound impact of artificial intelligence on our daily lives is exemplified by techniques such as chatbots and speech recognition. Leveraging these techniques (i.e., Chatbot and speech recognition), the paper aims to create a user-friendly mobile app to enhance English-speaking proficiency at minimal cost. This study employs three conversational AI models- ChatGPT, Blenderbot, and DialoGPT as chatbots, and uses Google's speech-to-text/text-to-speech models for voice recognition and generation. The developed app and the used models are evaluated through questionnaires and response time measurements, highlighting user satisfaction with the system's realism, accuracy, and user-friendliness. ChatGPT and Blenderbot emerge as top performers in response quality, while DialoGPT excels in response speed.

Downloads

Download data is not yet available.

Downloads

Published

2025-03-11

Issue

Section

Articles