多特征融合和机器学习理论的英语机考自动评分系统  

Multi⁃feature fusion and machine learning theory of English computer test automatic scoring system

作　　者：李发娟 LI Fajuan(School of Electronic Information and Computer Engineering,Sichuan Institute of Industrial Technology,Deyang 618500,China)

机构地区：[1]四川工业科技学院电子信息与计算机工程学院,四川德阳618500

出　　处：《电子设计工程》2025年第5期51-54,共4页Electronic Design Engineering

摘　　要：为更好地解决英语机考的差异评分问题,设计多特征融合和机器学习理论下的英语机考自动评分系统。构建自动评分模型,联合各级功能模块,完善自动评分方法,根据评分因素的多特征融合条件,定义机器学习算法模型,结合机器学习中的逻辑回归算法,设计出了科学合理的自动评分逻辑,实现对自动评分功能的需求分析。实验结果表明,所设计的自动评分系统可以有效消除评分者的主观差异,有助于保障英语机考评分过程的一致性。In order to better solve the problem of different scoring in English computer test,an automatic scoring system for English computer test based on multi-feature fusion and machine learning theory is designed.The automatic scoring model is constructed,combined with functional modules at all levels,and the automatic scoring method is improved.According to the multi-feature fusion conditions of scoring factors,the machine learning algorithm model is defined,and combined with the logistic regression algorithm in machine learning,a scientific and reasonable automatic scoring logic is designed to realize the demand analysis of automatic scoring function.The experimental results show that the designed automatic scoring system can effectively eliminate the subjective differences of raters and help to ensure the consistency of the scoring process of English computer test.

关 键 词：多特征融合 机器学习 自动评分 功能模块 功能需求 

分 类 号：TN876.3[电子电信—信息与通信工程]