摘要
设计了一种基于依存关系与同义词词林相结合的语义相似度计算方法。该方法通过依存关系分别提取两个文本的关系路径,同时基于同义词词林计算两个文本之间关系路径的语义相似度。在计算两个文本之间的语义相似度时,使用语言技术平台(language technology platform,LTP)对文本进行中文分词以及获取文本的依存关系图,从中提取关系路径,从而可以结合关系路径和同义词词林计算两个文本之间的语义相似度。通过实验,获得的平均偏差率为13.83%。实验结果表明,结合依存关系与同义词词林的语义相似度方法在准确率上相比较基于同义词词林的语义相似度和基于依存关系的语义相似度有了一定的提高。
We present a method of calculating semantic similarity based on the combination of dependency relation and Tongyici Cilin.This method extracts the relationship paths of two texts by the dependency relation,and calculates the semantic similarity of the relationship paths between two texts based on Tongyici Cilin.When calculating the semantic similarity between two texts,we use language technology platform(LTP)to segment the Chinese text and obtain the dependency graph of the text,and extract the relationship path from it,so that we can calculate the semantic similarity between the two texts by combining the relationship path and Tongyici Cilin.The average deviation rate is 13.83%in the experiment which shows that the accuracy of the semantic similarity method based on the dependency relation and Tongyici Cilin is better than that based on Tongyici Cilin and based on the dependency relation.
作者
付鹏斌
陈帅帅
杨惠荣
李建君
FU Peng-bin;CHEN Shuai-shuai;YANG Hui-rong;LI Jian-jun(Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China)
出处
《计算机技术与发展》
2020年第1期13-18,共6页
Computer Technology and Development
基金
北京市自然科学基金资助项目(4153058)
关键词
依存关系
同义词词林
语义相似度
关系路径
平均偏差率
dependency relation
Tongyici Cilin
semantic similarity
relationship path
average deviation rate
