基于分词匹配的主观题自动评阅技术研究  

Research on Subjective Question Automatic Marking Technology Based on Word Segmentation Matching

作　　者：冯光[1] 乔丹丹[1] 常静怡[1] 

机构地区：[1]安阳师范学院教育信息技术与传媒学院,河南安阳455000

出　　处：《计算机与现代化》2013年第3期212-214,219,共4页Computer and Modernization

基　　金：安阳师范学院大学生创新基金资助项目(ASCX/2011-Z22)

摘　　要：目前,计算机自动阅卷系统中,针对简答题、论述题等主观题的自动评阅技术发展尚不成熟。为了解决主观题自动评阅系统中存在的准确度问题,本文采用基于最大正向匹配的中文分词提取技术和关键词匹配技术实现一种性能稳定且准确率较高的主观题自动评阅系统。实验测试表明,本文的主观试题自动评分方法基本可以满足系统设计的需求,是有效可行的方法。Nowadays, because of the immaturity of the automatic marking system technology, some problems still exist when deal- ing with the subjective problems, such as short-answer questions, essay questions, etc. In order to improve the accuracy grade of assessing the subjective question automatically, this paper presents a kind of automated marking system with stable performance and high accuracy by adopting Chinese word segmentation extraction technology and key word matching technology which are based on forward maximum matching algorithm. Experimental results show that using this kind of automated marking system can satisfy the demand of system design and it is very effective and feasible.

关 键 词：中文分词 最大正向匹配 主观题 关键字匹配 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]