基于评论情感分析的个性化推荐策略研究——以豆瓣影评为例

发布时间：2018-04-30 03:30

本文选题：评论挖掘 + 情感分析　；参考：《情报理论与实践》2017年08期

【摘要】：[目的/意义]随着社会化媒体的兴起,信息资源的数量呈现爆炸式增长,如何在海量的信息中帮助用户发现有用的知识成为亟须解决的问题。互联网上已经存在的各类用户评论信息中蕴含着大量的可再开发的知识资源,包括用户的个人信息、选择偏好和消费习惯等,有助于解决"信息过载"问题。[方法/过程]文章通过对豆瓣电影评论信息进行细粒度的情感分析进而有效地获取集体智慧,并且利用评论挖掘技术发掘用户的偏好,为用户选择产品提供更加有效的推荐策略。[结果/结论]实验表明,将大众智慧与个性化服务两者有机地结合起来,能够真实地反映出不同用户对于电影的感受特性,并为用户观影提供更加合理的参考。
[Abstract]:[Objective / significance] with the rise of social media, the number of information resources presents an explosive growth. How to help users find useful knowledge in the mass of information has become an urgent problem. A large number of redeveloped knowledge resources are contained in all kinds of user comments on the Internet, including the individual's individual. Information, choice of preference and consumption habits help solve the "information overload" problem. [method / process] article can effectively obtain collective wisdom through a fine grained analysis of the comment information of the bean film, and use the review mining technology to discover the user's preference and provide more effective recommendation strategy for the user to choose the product. [results / Conclusion] the experiment shows that the combination of mass wisdom and personalized service can truly reflect the feelings of different users on the film, and provide a more reasonable reference for the user's view.

【作者单位】：南京大学信息管理学院;江苏省数据工程与知识服务重点实验室;西南大学图书馆;
【分类号】：G203

【相似文献】