Daliri Khomami Mohammad Mehdi, Rezvanian Alireza, Meybodi Mohammad Reza
Soft Computing Laboratory, Computer Engineering Department, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran.
Department of Computer Engineering, University of Science and Culture, Tehran, Iran.
Sci Rep. 2025 May 10;15(1):16322. doi: 10.1038/s41598-025-99948-z.
Identifying the most influential spreaders as an influence maximization problem (IMP) has become one of the most compelling topics in social network analysis due to its successes in viral marketing. In this paper, we first assume the network model to be a multiplex network, consisting of layers where each layer represents a different type of association among users based on their activities. We then define the concept of the maximum independent set (MIS) problem within multiplex networks. Next, we propose MIS as a potential solution to the MIP for identifying the initial candidate set of spreaders. Finally, we develop a learning automaton framework to solve the MIS in multiplex networks and to demonstrate its applicability for influence maximization. Theoretical properties of the MIS in multiplex networks are provided, along with various experiments on both artificial and real networks to showcase the performance of the proposed algorithm.
将识别最具影响力的传播者作为一个影响力最大化问题（IMP），由于其在病毒式营销中的成功应用，已成为社交网络分析中最引人注目的话题之一。在本文中，我们首先假设网络模型为一个多层网络，由多个层组成，其中每层基于用户活动代表用户之间不同类型的关联。然后，我们定义了多层网络中最大独立集（MIS）问题的概念。接下来，我们提出将MIS作为解决MIP以识别传播者初始候选集的潜在解决方案。最后，我们开发了一个学习自动机框架来解决多层网络中的MIS，并证明其在影响力最大化方面的适用性。文中给出了多层网络中MIS的理论性质，以及在人工网络和真实网络上进行的各种实验，以展示所提算法的性能。
