摘要
网络在线考试以其高效性和公正性正逐步取代传统的考试方法,客观题评阅通过将标准答案和学生答案进行精确比较给出评分,而主观题由于其复杂性和多样性成为机器阅卷算法关键点和难点。简单的关键词匹配算法速度较快但准确率不高,本文采用向量空间模型来表示文本,通过计算文本之间的相似度来进行主观题评阅,并在系统中增加一些策略和规则,取得了较好的阅卷效果。
Network online examination is gradually replacing traditional methods for its efficiency and impartiality. Objective question score can be obtained by using machine marking algorithm comparing standard answer with st-udent answer, while subjective question because of its complexity and diversi- - ty has become the key point and difficulty of machine marking algorithm. Simple keyword matching algorithm is fast, but accuracy is not high. Vector Space Model is used to represent text in this paper, and subjective question marking is based on the text similarity computation. Some strategies and rules are added into the system, and better marking result has been obtained.
出处
《安徽建筑工业学院学报（自然科学版）》
2010年第4期77-80,共4页
Journal of Anhui Institute of Architecture(Natural Science)
基金
安徽省高校优秀青年人才基金项目(2009SQRZ101)
关键词
相似度计算
向量空间模型
中文分词
similarity computation
Vector Space Model
Chinese word segmentation
作者简介
秦学勇（1974-），男，硕士，讲师，主要研究方向为智能信息检索。
