基于序列动态选择的口语机考自动评分系统设计

Design of automatic score system for spoken English test based on sequence dynamic selection

为了提高英语口语机考自动评分的自动化水平,提出一种基于序列动态选择的英语口语机考自动评分系统的设计方法,采用口语语音特征提取方法进行英语口语机考自动化评分的特征提取,分析英语口语机考评分的序列动态特征量,结合序列动态选择方法进行英语口语机考的语音相关性特征分析,建立英语口语机考自动特征匹配模型,采用块匹配和模板匹配方法,实现英语口语机考自动评分的序列相关性选择和关联规则挖掘,实现英语口语机考的自动化评分控制,在多层B/S构架体系下实现英语口语机考自动评分系统的软件开发设计。仿真结果表明,采用该方法进行英语口语机考自动评分系统设计的序列匹配性较好,具有很好的英语口语机考自动评分智能性。

In order to improve the automation level of automatic score in spoken English machine test,a design method of automatic scoring system based on sequence dynamic selection is proposed.The automatic score feature extraction of spoken English machine test is carried out by means of speech feature extraction method,and the sequence dynamic feature quantity of spoken English machine test score is analyzed.Based on the dynamic sequence selection method,the phonetic correlation features of spoken English machine test are analyzed,and the feature matching model of spoken English machine test is established,and the block matching and template matching methods are adopted.To realize the sequence correlation selection and association rule mining of the automatic score of spoken English machine test,and to realize the automatic score control of English spoken English machine test.The software development and design of the automatic score system of spoken English machine test is realized under the multi-tier B/S architecture.The simulation results show that the design of the automatic scoring system for spoken English machine test by this method has a good sequence matching performance and has a good intelligence of automatic scoring system of spoken English machine test.