Implemention of a Machine Intelligence Pronunciation System Based on Mul Tran to Improve the Basic English Speaking Skills

In the process of learning English, English pronunciation has always been difficult. This article was based on the Mul Tran (Multilingual Translation) platform and used various methods such as animation, sound, images, and text to learn and train English phonetics. It can provide effective pronunciation feedback for learners, guide and correct their continuous training, and improve their oral pronunciation. The system includes pronunciation demonstration, pronunciation follow-up, pronunciation comparison, and pronunciation scoring functions. The system also provides the function of “continuous following of original sounds”, which can continuously follow and score the original sounds, and comprehensively compare the pronunciation of each speech of the learners. For each pronunciation, according to the resonance peak image, the mouth shape was improved. The effective rate of correcting vowel pronunciation was 0.899, and the effective rate of correcting word pronunciation was 0.928, which had a certain guiding effect on student pronunciation correction.