| 摘 要: 当前，微博已经成长为世界上最有影响力的社交网络服务之一。随着微博的流行，微博上大量的数据也使 得用户无法快速获取他感兴趣的信息。推荐系统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协同过滤推荐技术的微博推荐算法，从影响用户兴趣度的隐性因 素，以及微博互联网中的数据采集和预处理等角度对微博推荐进行研究。使用矩阵分解对隐性因素建模，在已有用户与 微博、用户与微博发布者影响因素的基础上，提出微博与微博发布者影响因素，提高了原算法的准确度。 |
| 关键词: 微博推荐；协同过滤；矩阵分解 |
|
中图分类号: TP391
文献标识码: A
|
|
| A Personalized Micro-Blog Recommendation Algorithm Based on Collaborative Filtering |
|
QIN Xiaohui
|
( School of Computer Engineer, Taiyuan Institute of Technology, Taiyuan 030008, China)
|
| Abstract: Currently,micro-blog has become one of the most influential networking services throughout the world. Along with its increasing growth of popularity,the large number of information available on micro-blog has obstructed people from accessing the messages they are interested in.The micro-blog recommendation system picks out and recommends the objects (e.g.products,webpages,micro-blogs,etc.) via analyzing the existing data of the user.The paper proposes a microblog recommendation algorithm based on the collaborative filtering technique,explores some recessive factors which may influence user's interest and studies micro-blog recommendation from the perspective of data collecting and preprocessing on micro-blog networks.While the previous studies only focus on the relationship between the user and the publisher,and that between the user and the micro-blog post,this paper adopts matrix decomposition to model recessive factors and proposes the influence factors between the publisher and the micro-blog post.Finally,the experimental results show that the new algorithm significantly improves the accuracy of micro-blog recommendation. |
| Keywords: micro-blog recommendation;collaborative filtering;matrix decomposition |
