Analysis of English Oral Speech Information Based on Intelligent Algorithms

Abstract

In order to improve the intelligent evaluation effect of spoken English, this paper analyzes the traditional spoken English test algorithm, and proposes an improved spoken English scoring algorithm based on the needs of intelligent English evaluation.Moreover, this paper proposes a framework of multi-index fusion pronunciation quality evaluation technology for reading questions, and constructs a functional module for English oral proficiency evaluation based on the needs of speech feature recognition. At the same time, this paper uses deep learning-based speech recognition technology to automatically recognize the tester’s pronunciation, and uses the dual-threshold endpoint detection method of short-term energy and short-term average zero-crossing rate to divide the pronunciation of the speech sentence into syllables.In addition, this paper recognizes speech features by inputting the candidates’ voice, and compares the recognition features with the standard database to score it. Finally, this paper analyzes the system performance by means of experimental research. The research results show that the system constructed in this paper has a certain effect.