Design of An Automatic Quality Evaluation System for English Pronunciation Machines Based on SVM

With the acceleration of the process of global economic integration, English has become the main way of communication in the international community today. The evaluation of the voice quality of spoken English is a very important part of spoken English teaching. The article expounds the conditions for evaluating the quality of English speech. Based on the acquisition of English speech signals, it is combined with the Support Vector Machine (SVM) method to study the characteristics of English speech signals. This article introduced an automatic English voice quality evaluation system. In terms of experiments, the first half of the system was given, and the collected English voice signal data was evaluated accordingly. Through the analysis of experimental data, the accuracy of the English speech quality evaluation system was verified. The experimental results showed that the classification effect of support vector machines was better than that of traditional classification algorithms. The automatic evaluation system of English speech quality established by the support vector machine method, the evaluation results increase with the increase of the number of evaluations, and the error of the evaluation results was small, all within 0.2. The maximum error value of the automated evaluation of linear regression was greater than 0.6. Through the analysis of English speech, the correctness of this method is verified, and the SVM method is used to evaluate the method more accurately. Therefore, the SVM method can improve the effect of English speech evaluation.