基于有向图模型的评论方面挖掘研究
首发时间:2018-01-12
摘要:评论的方面挖掘为商品属性提供一个直观的概要,对电子商务的发展具有重要意义。本文提出了一种基于依存关系与有向图模型的评论对象与评价词的提取方法,利用依存语法分析,构建以词为结点,以词间直接与一阶间接依存度为边权重的有向图模型;提取相关评价对象、评价词二元组,按边权重进行排序筛选;通过对评价对象进行聚类、分类来进行语义去重,最后利用评价对象与评价词进行联合情感分析。基于SemEval 2015测评数据进行仿真实验,实验结果表明本文所提出的基于有向图模型的方面挖掘算法可以有效的完成评价对象的提取以及评价情感的分析。
For information in English, please click here
Research on aspect mining of reviews based on directed graph model
Abstract:The aspect mining of comments provides an intuitive overview of product attributes and is of great importance to the development of e-commerce.In this paper, a method for extracting comment aspects and opinions based on grammar dependency relationship and directed graph model is proposed. Based on the grammar dependency relationship analysis, the graph is constructed with words as nodes, the direct and the first order indirect dependency as the weight of the edges.Extract the tuple of aspect and opinion, then selectthe tuples sorted by the weight of the edges of the graph. Semantically deduplicate the aspects by clustering and classification algorithm and analysis the sentiment of the comment by the aspects and opinions jointly. The simulation experiment is based on the data of SemEval 2015, the results show that the proposed aspect mining algorithm baed on directed graph model can effectively extract the aspects and opinions of the comments.
Keywords: artificial intelligence aspect mining directed graph grammar dependency
基金:
引用
No.****
同行评议
勘误表
基于有向图模型的评论方面挖掘研究
评论
全部评论0/1000