An Intelligent Evaluation Model of English Pronunciation Quality Based on Sphinx

摘要

Based on the preliminary study of speech recognition technology, scoring methods and relevant theoretical knowledge of information feedback, this paper proposes an improved scheme using posterior probability scoring, and establishes an expert opinion database, according to the common types of oral pronunciation errors. Combining with the corrective information of the database, it can effectively help learners improve their oral proficiency. Then, we design an automatic scoring system for spoken English. It realizes the functions of recognition, scoring, error determination and feedback correction of user's pronunciation based on speech recognition device. Finally, through experiments and tests, it is proved that the system can effectively improve the scoring performance of the system after introducing the average pronunciation level, and it has a certain effect on improving the spoken pronunciation level of users.