摘要
针对大规模考试管理中存在相似试卷识别的困难,提出一种试卷相似度自动评估算法。参考现有知网词汇语义相似度计算方法,结合试卷相似度计算领域特点,改进词汇语义相似度计算方法,提出试题相似度和试卷相似度计算模型,实现对试卷相似度的自动评估,提高了大规模考试管理的工作效率。通过抽取各专业部分试卷进行相似度自动评估测试,测试结果与人工评估结果基本一致,达到预期目标。
For the difficulty of similar test papers' evaluation in large-scale examination management,an automatic evaluation algorithm is presented in the paper.Making reference to the existing calculation method of word semantic similarity based on HowNet,and combining with the characteristics of test paper similarity,the paper improves the method of semantic similarity calculation,and proposes the similarity model of test questions and test papers,which achieves the automatic evaluation of test paper's similarity,and improves the efficiency of large-scale examination management.By extracting test papers from various professions to take test,the test results are basically consistent with the manual assessment results,and achieves the desired goal.
作者
郑炜冬
ZHENG Weidong(Education Department;Hanshan Normal University;Chaozhou Guangdong 521041;China)
出处
《智能计算机与应用》
2011年第4期16-19,共4页
Intelligent Computer and Applications
基金
韩山师范学院教学改革研究项目“《现代教育技术》考试平台开发与应用”(HJG1107)
关键词
试卷相似度
自动评估
语义相似度
知网
Paper's Similarity
Automatic Evaluation
Semantic Similarity
HowNet