摘要
主观题阅卷系统在目前较受人们的关注。主观题评分原则常见的是对比试卷答案与参考答案的相似程度,进而得出分数。利用自然语言处理分词技术将文本切分为词的组合,然后通过改进语义相似度算法,计算句子相似度,进而得到试卷答案和参考答案的相似程度,给出分数。
Subjective item marking system has been a study focus.The common method is to contrast the answers with the reference answers to form a score.This paper uses the technology of natural language processing participle to divide a sentence into the assemble of phrases,and then get the score by computing a sentence similarity degree with improved semantic similarity algorithm.
出处
《河北科技大学学报》
CAS
2012年第3期263-265,共3页
Journal of Hebei University of Science and Technology
关键词
主观题
分词
语义相似度
subjective item
participle
semantic similarity
