Exploiting Popularity and Similarity for Link Recommendation in Twitter Networks
Jun Zou, Faramarz Fekri · 2014
Twitter functions both as a social network and an informa-tion network, where users follow other users to make social connections as well as to receive information. Both pop-ularity and similarity are important factors that drive the growth of the Twitter network. In this paper, we propose two approaches to exploiting both popularity and similarity for link recommendation. The first approach employs the rank aggregation technique to combine rankings generated by popularity-based and similarity-based recommendation algorithms. The second approach adapts the collaborative filtering algorithms to incorporate popularity in addition to similarity. The empirical evaluation results on real-world datasets confirm that combining popularity and similarity improves the recommendation performance.