社会网中基于主题的局部影响最大化算法研究.pdfVIP

社会网中基于主题的局部影响最大化算法研究* 谢胜男，刘 勇+，朱敬华，谭 龙 黑龙江大学 计算机科学与技术学院，哈尔滨 150080 Research on Topic-Based Local Influence Maximization Algorithm in Social Network􀆽 XIE Shengnan, LIU Yong+, ZHU Jinghua, TAN Long School of Computer Science and Technology, Heilongjiang University, Harbin 150080, China + Corresponding author: E-mail: acliuyong@ XIE Shengnan, LIU Yong, ZHU Jinghua, et al. Research on topic-based local influence maximization algo- rithm in social network. Journal of Frontiers of Computer Science and Technology, 2016, 10(5): 646-656. Abstract: Local influence maximization is the problem of finding a small set of seed nodes in a social network that maximizes the influence on a target node. Existing works only consider the influence on a single target node, and also ignore the topic distribution of diffusion items and the influence probabilities between users based on topics. This paper focuses on how to select a seed set that has the maximum influence on a given target set based on the topic distri- bution, and proposes a new topic-based local influence degree computational method (T-LID). Based on T-LID, this paper also proposes a topic-based local influence maximization (TLIM) problem and proves that TLIM is NP-hard. In order to solve TLIM, this paper proposes a topic-based local greedy algorithm (TLGA) and a topic-based local propa- gation algorithm (TLPA). The experimental results on several real datasets show that the proposed algorithms can solve the TLIM problem effectively and efficiently. Key words: social network; topic-based; local influence maximization; target set 摘 要：局部影响最大化问题是在社会网中寻找最能影响某个目标节点的种集。现有的研究只考虑对单一目 标节点的影响，忽略了传播项上的主题分布以及用户之间基于主题的影响概率。重点研究了在主题分布的条 * The National Natural Science Foundation of China under Grant No.(国家自然科学基金); the Natural Science Foundation of Heilongjiang Province under Grant No. F201430 (黑龙江省自然科学基金); the Scientific Research Fund of Heilongjiang Provin- cial Education Department under Grant Nos.(黑龙江省教育厅科