摘要
句子相似度的计算在自然语言处理的各个领域都占有很重要的地位,且在主观题自动批改技术中,句子间相似度的计算是一个关键的问题.本文提出一种基于多层次融合的语句相似度计算模型,在对句子的词形、词序、词义三个层次进行相似度计算的基础上,融合这三个层次,从而使句子相似度计算更加准确,并通过实验验证了此方法的有效性.
The sentence-similarity holds the very important status in natural language disposal. And computing sentence-similarity between sentences is the key problem in marking subjective topic automatically. A model based on the multi-level sentence-similarity is proposed in this paper, to fuse those three levels which based on the sentence morphology, the word order, and the word meaning, so that the sentence-similarity to be more accurate, the reliability of the method is validated by the results of experiment.
出处
《延边大学学报（自然科学版）》
CAS
2007年第3期191-194,共4页
Journal of Yanbian University（Natural Science Edition）
关键词
语句相似度
自然语言处理
多层次融合
知网
sentence-similarity
natural language processing
multi-level fusion
How-net
