| 在线用户评论的主题发现研究 |
| |
| 引用本文： | 王和勇,崔蓉.在线用户评论的主题发现研究[J].现代情报,2015,35(9):63-69. |
| |
| 作者姓名： | 王和勇 崔蓉 |
| |
| 作者单位： | 华南理工大学电子商务系, 广东 广州 510006 |
| |
| 摘 要： | 在线用户评论是电子商务网站中的一个重要板块,找出在线用户评论的关注点有利于网站、商家及时有效地查看用户的反馈信息。本文在对在线用户评论进行分词的基础上,分别使用拉普拉斯评分(LS,Laplacian Score)及信息增益(IG,Information Gain)对所得到的分词结果进行文本主题挖掘,并使用支持向量机(SVM,Support Vector Machine)进行分类精度的检验。实证结果表明,主题选择的结果是有效的,分类的效果与选择的关键词个数和核函数有关。
|
Research on Online User Reviews Topic Discovery |
| |
| Authors: | Wang Heyong Cui Rong |
| |
| Institution: | Department of E-Business, South China University of Technology, Guangzhou 510006, China |
| |
| Abstract: | Online users' reviews are important for e-business website, and finding the topic of these reviews can help both websites and businesses pay close attention to users' feedbacks.Based on word segmentation, this paper separately used laplacian score(LS) and information gain(IG) to find text topic, and then support vector machine(SVM) method was used to verify the classification accuracy.The results showed that it is effective to select the text topic and the accuracy is related to the number of topic and the core function used. |
| |
| Keywords: | |
|
| 点击此处可从《现代情报》浏览原始摘要信息 |
| 点击此处可从《现代情报》下载免费的PDF全文 |
|
