Improved Algorithms OF CELF and CELF++ for Influence Maximization

Motivated by the wide application in some fields, such as viral marketing, sales promotion etc, influence maximization has been the most important and extensively studied problem in social network. However, the most classical KK-Greedy algorithm for influence maximization is inefficient. Two major sources of the algorithm’s inefficiency were analyzed in this paper. With the analysis of algorithms CELF and CELF++, all nodes in the influenced set of u would never bring any marginal gain when a new seed u was produced. Through this optimization strategy, a lot of redundant nodes will be removed from the candidate nodes. Basing on the strategy, two improved algorithms of Lv_CELF and Lv_CELF++ were proposed in this study. To evaluate the two algorithms, the two algorithms with their benchmark algorithms of CELF and CELF++ were conducted on some real world datasets. To estimate the algorithms, influence degree and running time were employed to measure the performance and efficiency respectively. Experimental results showed that, compared with benchmark algorithms of CELF and CELF++, matching effects and higher efficiency were achieved by the new algorithms Lv_CELF and Lv_CELF++. Solutions with the proposed optimization strategy can be useful for the decisionmaking problems under the scenarios related to the influence maximization problem.

[1]  Masahiro Kimura,et al.  Efficient discovery of influential nodes for SIS models in social networks , 2011, Knowledge and Information Systems.

[2]  Y. Narahari,et al.  A Shapley Value-Based Approach to Discover Influential Nodes in Social Networks , 2011, IEEE Transactions on Automation Science and Engineering.

[3]  Suh-Yin Lee,et al.  Efficient algorithms for influence maximization in social networks , 2012, Knowledge and Information Systems.

[4]  Wan-Shiou Yang,et al.  Application of the Ant Colony Optimization Algorithm to the Influence-Maximization Problem. , 2012 .

[5]  Matthew Richardson,et al.  Mining the network value of customers , 2001, KDD '01.

[6]  Wei Chen,et al.  Efficient influence maximization in social networks , 2009, KDD.

[7]  Weili Wu,et al.  The complexity of influence maximization problem in the deterministic linear threshold model , 2012, J. Comb. Optim..

[8]  Tian Jia,et al.  A New Hybrid Algorithm for Influence Maximization in Social Networks , 2011 .

[9]  Wei Chen,et al.  Scalable influence maximization for independent cascade model in large-scale social networks , 2012, Data Mining and Knowledge Discovery.

[10]  Laks V. S. Lakshmanan,et al.  On minimizing budget and time in influence propagation over social networks , 2012, Social Network Analysis and Mining.

[11]  Éva Tardos,et al.  Maximizing the Spread of Influence through a Social Network , 2015, Theory Comput..

[12]  Wei Chen,et al.  Scalable influence maximization for prevalent viral marketing in large-scale social networks , 2010, KDD.

[13]  Matthew Richardson,et al.  Mining knowledge-sharing sites for viral marketing , 2002, KDD.

[14]  Andreas Krause,et al.  Cost-effective outbreak detection in networks , 2007, KDD '07.

[15]  Laks V. S. Lakshmanan,et al.  CELF++: optimizing the greedy algorithm for influence maximization in social networks , 2011, WWW.

[16]  Jingyu Zhou,et al.  Preference-based mining of top-K influential nodes in social networks , 2014, Future Gener. Comput. Syst..

[17]  Noga Alon,et al.  Information Processing Letters a Note on Competitive Diffusion through Social Networks , 2022 .

[18]  Yifei Yuan,et al.  Scalable Influence Maximization in Social Networks under the Linear Threshold Model , 2010, 2010 IEEE International Conference on Data Mining.