摘要
为了实现考试系统中主观题型的自动阅卷,提出了基于语句相似度的计算方法对主观题进行自动批改。通过对语句相似度的定义和计算模型的建立以及汉语的特殊性分析,同时借鉴自动问答、自动文摘、信息检索等领域内关于语句相似度的研究,建立了对主观题的自动阅卷模型。通过对语句的词形、词序、词义3个层次计算相似度的融合,使句子的相似度计算更加准确。实验结果表明,主观题自动阅卷技术的正确率达到81.5%。
A new method on automated assessment technology for subjective tests is proposed based on sentence similarity,which can be used in exam system to estimate subjective tests such as essay question.The definition and the model based on sentence similarity achieved action and Chinese particularity is analyzed,at the same time many techniques such as QA system,automatic summarization and information retrieval etc.are used to study the sentence similarity,set up a model on automated assessment for subjective tests.On the basis of evaluating the similarity according to the sentences’ morphology,semantic and acceptation,amalgamating these three levels,it is more accurate to evaluate the sentence similarity.By specific experimental tests with the method above,success rate attained 81.5%.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第16期3697-3699,3704,共4页
Computer Engineering and Design
关键词
自然语言处理
自动阅卷
主观题
相似度计算
知网
natural language processing
automated assessment
subjective tests
similarity computation
HowNet
