Answers are grounded in this paper and cite their sources.
Different from conventional social networks and e-commerce systems, microblog community is unique for its low user activity, data sparsity and dynamic of user’s interests. Because of these challenges, conventional recommendation algorithms cannot get a satisfactory performance in microblog community. This paper proposes a microblog followee recommendation algorithm based on user interest degree and attribute characteristics after analyzing the structure of microblog recommendation, and this algorithm is mainly on the basis of user cf recommendation algorithm to recommend new followee to user. Firstly, group users through calculating users interest degree to followee; Secondly, get users’ preferences value to followee with Bayesian algorithm when they are with different attribute characteristics; Finally, calculate target user’s nearest neighbor set using a new optimized similarity degree method to form the recommendation list. Experiments show that this proposed algorithm enhances the effectiveness and accuracy of the nearest neighbor set, and the recommendation quality has significantly improvement compared with previous algorithm.
No takes yet. Share an insight, caveat, or question.
Rongsheng Wang (2014) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: