ADVANCING COMMUNICATION IN SPORTS COACHING: AN AI-BASED METHOD FOR EVALUATING ORAL ENGLISH PRONUNCIATION

Abstract

Objective: This study develops an artificial intelligence-based method for evaluating oral English pronunciation, aiming to enhance communication effectiveness in sports coaching and training environments. Traditional pronunciation evaluation methods often rely on fixed standards and templates, leading to limited accuracy in real-world applications. Methods: The proposed method utilizes a Hidden Markov Model (HMM) to preprocess and establish a spoken pronunciation acoustic model from oral English pronunciation signals. By analyzing spoken speech characteristics from a phonemic perspective and extracting key pronunciation features, the system employs a neural network to derive accurate evaluation results. Results: Testing of this AI-driven evaluation method demonstrated an accuracy rate exceeding 92%, with a notably low misdetection rate, indicating significant improvements over traditional methods. The enhanced accuracy and reliability suggest that this method can effectively support coaches and trainers in assessing and improving English pronunciation among athletes, facilitating clearer and more effective communication. Conclusion: The AI-based oral English pronunciation evaluation method offers substantial potential for application in sports settings, where clear and precise communication is essential for successful coaching and athlete performance. This approach not only improves the quality of pronunciation assessment but also contributes to more effective training and interaction in multilingual sports environments. Future research should explore the integration of this technology into diverse coaching tools and platforms, enhancing accessibility and usability across different sports disciplines.