随着在线社会网络的发展以及个性化信息推荐服务的应用,基于社会网络的协同推荐技术得到了发展。将社会网络的关键用户融入到传统的基于用户聚类的协同过滤算法,即通过关键用户改进协同过滤算法,并通过实验验证关键用户对协同过滤算法性能的影响。实验结果表明关键用户可以提高协同过滤算法的推荐准确性,这也表明社会网络中用户兴趣与关键用户的兴趣是相关的。%With the development of online social network and the application of personalized recommendation service, collaborative rec-ommendation techniques based on social network have been developed. By integrating key users of social network into the traditional user-based clustering collaborative filtering ( CF) algorithm, this paper uses key users to enhance CF algorithm and verifies the influence of key users on the performance of CF algorithm by experiments. Experimental results show that the key users can improve the accuracy of CF algorithm, which also shows that user interests are related with the interests of key users in social network.
展开▼