| 摘 要: 随着社交媒体的发展，微博为人们提供的服务正在极大地改变着人们使用互联网的习惯，然而微博上用 户发表的大量信息，以及高频率的信息更新，使得用户面临信息过载的问题而无法快速获取他感兴趣的信息。推荐系统 是解决此问题的一种很好的方法，它是通过研究用户已有数据来发掘用户兴趣，从而为用户推荐可能感兴趣的对象，如 产品、网页、微博等。本文介绍了一种基于张量分解技术的微博推荐算法来预测用户对微博的兴趣度，同时考虑用户与 微博、用户与微博发布者影响因素，以及微博与微博发布者的影响因素，提高了已有算法的准确度。 |
| 关键词: 微博推荐 矩阵分解 张量分解 |
|
中图分类号: TP391
文献标识码: A
|
| 基金项目: 太原工业学院科学基金项目(2016LQ14). |
|
| A Study of the Personalized Micro-Blog Recommendation Algorithm Based on Tensor Factorization |
|
QIN Xiaohui
|
( School of Computer Engineer, Taiyuan Institute of Technology, Taiyuan 030008, China)
|
| Abstract: With the development of social media,the services in micro-blog have significantly changed the way people use the Internet.However,as the large amount of information posted by users and the highly frequent update on micro-blogs,users often face the problem of information overload and miss out the content they are interested in.The recommendation system,which recommends items(such as products,web pages,micro-blogs,etc.)to users based on their interests,is an effective solution to this problem.The paper introduces a micro-blog recommendation algorithm based on the tensor factorization technology to predict the user's interest degree on certain micro-blog.The experimental results on real dataset show that the proposed model achieves desirable performance in characterizing the user's interest and the preprocessing of data on micro-blog.Finally,the paper presents the experimental results which show that the method significantly outperforms the baseline method. |
| Keywords: micro-blog recommendation matrix factorization tensor factorization |
