产品垃圾评论检测研究综述  

Survey on Product Review Spam Detection

作　　者：孙升芸[1] 田萱[1] 

机构地区：[1]北京林业大学信息学院,北京100083

出　　处：《计算机科学》2011年第B10期198-201,共4页Computer Science

基　　金：中央高校基本科研业务费专项资金(YX2011-30);北京林业大学新进教师科研启动基金(BLX2w8019)资助

摘　　要：互联网上的产品垃圾评论混淆视听,误导了潜在消费者。产品垃圾评论检测的目的就是将垃圾评论从评论文本中找到并去除,保留真实的产品评论供用户参考。首先将产品垃圾评论和互联网上其它常见的垃圾信息进行了对比,并把产品垃圾评论的检测和产品评论的质量判断、产品评论的情感分析等相关的工作进行了比较分析。然后从产品垃圾评论检测的数据集、检测方法两个角度对相关工作做了概述和分析。最后,在上述工作的基础上提出了一些产品垃圾评论检测研究中值得进一步关注的问题。It is common for e-commerce Web sites to enable their customers to write reviews of products that they have purchased.Unfortunately,reviewers may write some untruthful opinions in order to promote or damage specific pro-ducts＇ reputation which called review spam.Product review spam detection makes an attempt to find untruthful opinions and remove them to restore the truthful for readers.The difference between product review spam and other spams was analyzed firstly.The related work of review spam detection was presented including review helpfulness and review sentiment analysis later：and then the data set and methods to detect spam were introduced and summarized.The trend of product review spam detection was concluded finally.

关 键 词：产品评论 垃圾评论 垃圾评论检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]