Hybrid followee recommendation in microblogging systems  

作　　者：Hanhua CHEN Hai JIN Xiaolong CUI 

机构地区：[1]Services Computing Technology and System Laboratory, Big Data Technology and System Laboratory Cluster and Grid Computing Laboratory, School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, China

出　　处：《Science China(Information Sciences)》2017年第1期17-30,共14页中国科学（信息科学）（英文版）

基　　金：supported in part by National Natural Science Foundation of China (Grant Nos. 61370233, 61422202);Research Fund of Guangdong Province (Grant No. 2015B010131001);Foundation for the Author of National Excellent Doctoral Dissertation of China (Grant No. 201345)

摘　　要：Followee recommendation plays an important role in information sharing over microblogging platforms. Existing followee recommendation schemes adopt either content relevance or social information for followee ranking, suffering poor performance. Based on the observation that microblogging systems have dual roles of social network and news media platform, we propose a novel followee recommendation scheme that takes into account the information sources of both tweet contents and the social structures. We set up a linear weighted model to combine the two factors and further design a simulated annealing algorithm to automatically assign the weights of both factors in order to achieve an optimized combination of them. We conduct comprehensive experiments on real-world datasets collected from Sina Weibo, the largest microblogging system in China. The results demonstrate that our scheme provides a much more accurate followee recommendation for a user compared to existing schemes.Followee recommendation plays an important role in information sharing over microblogging platforms. Existing followee recommendation schemes adopt either content relevance or social information for followee ranking, suffering poor performance. Based on the observation that microblogging systems have dual roles of social network and news media platform, we propose a novel followee recommendation scheme that takes into account the information sources of both tweet contents and the social structures. We set up a linear weighted model to combine the two factors and further design a simulated annealing algorithm to automatically assign the weights of both factors in order to achieve an optimized combination of them. We conduct comprehensive experiments on real-world datasets collected from Sina Weibo, the largest microblogging system in China. The results demonstrate that our scheme provides a much more accurate followee recommendation for a user compared to existing schemes.

关 键 词：online social networks MICROBLOGGING followee recommendation simulated annealing RANKING 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]