English Pronunciation Evaluation and Correction Algorithm Technology Based on Artificial Intelligence

Based on artificial intelligence, this thesis studies evaluating and correcting English pronunciation. Traditional English teaching evaluation methods are subjective, labourconsuming, time-limited and lack pertinent teaching guidance, which seriously hinders the further improvement of students’ English pronunciation. This article aims to develop a new evaluation and correction method based on speech recognition, speech synthesis, machine learning and natural language processing. The system can objectively evaluate the students’ pronunciation, give them immediate feedback, and correct their mistakes individually. Through accurate assessment and targeted correction, students can better grasp the pronunciation of spoken English. After six months of pronunciation improvement training, one student’s pronunciation accuracy increased from 0.65 to 0.911 , while the others improved. The research results of this project can make up for the deficiencies of previous studies and have great theoretical significance and practical application value for English phonetics teaching.