基于用户兴趣的微博信息推荐研究

Weibo Information Recommendation Research Based on User Interest

微博作为信息获取、共享及推广的手段,受到大众的广泛喜爱。随着微博的流行,微博用户也面临着一种信息严重超负荷的局面,致使他们无法快速获取到感兴趣的信息。随着推荐系统的应用,使这种问题得到了有效解决。本文在研究的过程中,主要与微博用户兴趣度的隐性因素相结合,基于改进人工鱼群算法开发设计了协同过滤推荐模型,由实验结果可知,根据对比传统的协同过滤推荐算法可得之,该算法的优势明显,具备较佳的优化精度与收敛性。

Weibo,a useful tool for gaining,sharing and promotion of information,has received widespread popularity among people.While Weibo is becoming increasingly prevalent,Weibo users find themselves in face of a difficult situation of information overloading.As a consequence,they don’t have quick access to information which interests them.Application of the information recommendation system proves to be an effective solution to this problem.In this thesis,our research has developed and designed a collaborative filtering recommendation model by combining recessive factors of Weibo user interest and on the basis of improved artificial fish swarm algorithm.According to the experiment results,this new algorithm featuring better accuracy optimization and astringency has remarkable advantages over traditional collaborative filtering recommendation algorithm.