MATI: An efficient algorithm for influence maximization in social networks

被引:10
作者
Rossi, Maria-Evgenia G. [1 ]
Shi, Bowen [1 ]
Tziortziotis, Nikolaos [1 ]
Malliaros, Fragkiskos D. [2 ,3 ]
Giatsidis, Christos [1 ]
Vazirgiannis, Michalis [1 ]
机构
[1] Ecole Polytech, Palaiseau, France
[2] Univ Paris Saclay, Ctr Supelec, Gif Sur Yvette, France
[3] Inria Saclay, Gif Sur Yvette, France
关键词
IDENTIFICATION;
D O I
10.1371/journal.pone.0206318
中图分类号
O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];
学科分类号
07 ; 0710 ; 09 ;
摘要
Influence maximization has attracted a lot of attention due to its numerous applications, including diffusion of social movements, the spread of news, viral marketing and outbreak of diseases. The objective is to discover a group of users that are able to maximize the spread of influence across a network. The greedy algorithm gives a solution to the Influence Maximization problem while having a good approximation ratio. Nevertheless it does not scale well for large scale datasets. In this paper, we propose Matrix Influence, MATI, an efficient algorithm that can be used under both the Linear Threshold and Independent Cascade diffusion models. MATI is based on the precalculation of the influence by taking advantage of the simple paths in the node's neighborhood. An extensive empirical analysis has been performed on multiple real-world datasets showing that MATI has competitive performance when compared to other well-known algorithms with regards to running time and expected influence spread.
引用
收藏
页数:21
相关论文
共 44 条
[1]  
[Anonymous], DAT MIN ICDM 2011 IE
[2]  
[Anonymous], S JOHN SOCIAL NETWOR
[3]  
[Anonymous], ADV NEURAL INFORM PR
[4]  
[Anonymous], P 3 ACM INT C WEB SE
[5]  
[Anonymous], 2010, P 16 ACM SIGKDD INT
[6]  
[Anonymous], AAAI
[7]  
[Anonymous], AAAI SPRING S COMP A
[8]  
[Anonymous], P 23 ACM INT C C INF
[9]  
[Anonymous], 2006, P 29 ANN INT ACM SIG
[10]  
[Anonymous], P 2014 ACM SIGMOD IN