摘要
从自然语言处理的角度来看,现有的主观题评分算法都以相似度计算为核心的,没有考虑语义对立度问题.提出了一种基于中文分词技术、相似度计算和对立度计算的新的主观题评分算法.对主观题评分算法的以下三个方面进行了重点研究:怎样改进中文分词算法中的歧义切分的方法;怎样引入参数限制计算中分数过高或过低;怎样设计一个融合相似度计算和对立度计算的主观题分数的计算公式.
From the perspective of natural language processing, all existing subjective questions scoring algorithms use similarity calculation as a key method while the contrary degree is ignored. A new subjective questions scoring algorithm based on Chinese words segmentation technology, similarity calculation and contrary degree calculation is presented in this paper. The three areas about subjective questions scoring algorithm are mainly researched ： how to improve the method of disambiguation in Chinese words segmentation ; how to introduce parameters to restrict the score being too high or too low in calculation; and how to design a formula to combine similarity calculation and contrary degree calculation for subjective questions scoring.
出处
《昆明理工大学学报（理工版）》
北大核心
2010年第2期81-84,共4页
Journal of Kunming University of Science and Technology(Natural Science Edition)
基金
广西自然科学基金资助项目(项目编号:桂科自0991254)
关键词
主观题
相似度
对立度
subjective question
similarity
contrary degree
