论文部分内容阅读
网上购物已成为社会生活中的一部分,商品购买评论为消费者和厂商提供了重要的产品信息反馈。通过构建领域本体、改进情感词典并构造商品特征匹配组的方法,比较京东网、亚马逊商城、天猫商城三个购物网站的商品评论情感倾向。结果发现不同网站关于手机各部件特征的评论在数量上和情感倾向上差异明显,一定程度上反映出网站用户群体及商品版本之间的差异。通过比较多网站的商品评论,帮助消费者获得较单一网站更为可信的商品评论,并为厂商改进产品提供参考。
Online shopping has become a part of social life, and product purchase reviews provide important product feedback to consumers and vendors. Through the construction of domain ontology, the improvement of thesaurus of sentiment and the construction of the product feature matching group, we compared the sentiment tendencies of product reviews on three shopping websites: Jingdong.com, Amazon.com and Tmall. The results showed that there were significant differences in the number and emotional tendencies of different websites on the features of mobile phones, to a certain extent, reflecting the differences between the website user groups and product versions. By comparing product reviews from many websites, we can help consumers get more credible product reviews than a single website and provide references for manufacturers to improve their products.