基于隐含语义分析的电商虚假评论识别  

Identification of Fake Online Products' Reviews based on Latent Semantic Analysis

作　　者：李存林 杨世瀚[1] 王晗[1] LI Cun-lin;YANG Shi-han;WANG Han(Guangxi Key Laboratory Guangxi of Hybrid Computation and IC Design Analysis for Nationalities, Nanning 530006, Chin)

机构地区：[1]广西民族大学广西混杂计算与集成电路设计分析重点实验室,广西南宁530006

出　　处：《广西民族大学学报（自然科学版）》2018年第1期52-59,共8页Journal of Guangxi Minzu University ：Natural Science Edition

基　　金：国家自然科学基金(11371003;11461006);广西科技基地和人才专项(2016AD05050);广西"八桂学者"专项;广西自然科学基金(2014GXNSFAA118359)

摘　　要：针对目前虚假评论识别都是通过评论者或评论文本的显式特征来进行的,但是没有考虑评论文本的语义信息,分析结果不够准确,因此提出基于隐含语义分析的电商虚假评论识别方法.该方法在基于用户行为特征分析的基础上增加了评论文本的语义分析信息,能在文本语义层次上识别虚假评论,提升了识别准确性.方法中给出了评论文本的可信度度量并针对有序的可信度度量提出了一种新的错位级别评估方法.Aiming at fake reviews recognition is carried out through the explicit characteristics of critics or review texts,but did not consider the semantic information of text comments,the results are not accurate enough,so put forward the latent semantic analysis of fake reviews recognition method based on electricity supplier comments.Based on the analysis of user behavior characteristics,this method increases the semantic analysis information of the comment text,which makes it possible to identify the fake reviews on the semantic level of the text and improve the recognition accuracy.We propose a new method level of dislocation to evaluate the reliability of the comment text.

关 键 词：商品评论 隐含语义分析 虚假评论识别 错位级别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]