A Machine Learning Assessment System for Spoken English Based on Linear Predictive Coding

In the teaching of English, there is an increasing focus on practical communication skills. As a result, speaking test component has received more and attention from education experts. With rapid development modern computer technology network technology, use computers to assess quality spoken English become hot topic research in related fields at present. A machine learning assessment system based linear predictive coding proposed order achieve automatic scoring tests. First, principle decoding analyzed, traditional algorithm improved by using hybrid excitation instead binary excitation. Second, overall structure designed, which mainly includes division into four modules: acoustic model acquisition module, speech recognition standard pronunciation transcription decision module. Then, module implemented method acquire feature parameters signal generate vector. Finally, convolutional neural used train features so as implement The experimental results show that yields natural higher intelligibility signals. designed evaluation able accurately detect information about learner’s pronunciation.