摘要：Spoken English practice requires a combination of listening, speaking, reading, and writing, among which listening and speaking are the most difficult. In order to improve the speaking ability of the practitioner, the pronunciation of spoken English needs to be corrected in time. However, the workload of manual evaluation is too large, so it is necessary to combine intelligent methods for spoken language recognition. Based on the needs of spoken English pronunciation correction, this paper combines the computer English speech recognition technology to construct the spoken English recognition and correction model and combines the coding technology to study the English speech recognition technology. Moreover, this article constructs the spoken English practice system based on the actual needs of spoken English practice. Finally, this paper verifies the reliability of this system through experimental research, which provides a reliable means for the subsequent intelligent learning of spoken English.
