The detection of top influential users in social networks is considered one of the current vital research field. The spreading of the information in social networks can be analyzed and sometimes controlled by studying those top influential users. This paper proposes LKG, a fast and scalable hybrid approach to detect top influential users in social networks, suitable for both directed and undirected networks. The LKG hybrid approach consists of three phases: (1) community detection, in which the complete social network is partitioned into related communities using the Louvain algorithm; (2) detection of community top nodes by applying the k-shell decomposition locally in each portioned community; and (3) selection generalization, in which the prior obtained results are generalized for maximizing the spread of influence. Experimental studies were conducted on several datasets with different sizes. The results have been shown to achieve better results for the spread of influence using incomplete social networks than the existing related work models and with far much less processing time.

社交网络中最具影响力的用户的检测被认为是当前重要的研究领域之一。信息在社交网络中的传播可以通过研究那些最有影响力的用户来分析和控制。本文提出了 LKG，一种快速且可扩展的混合方法，用于检测社交网络中最有影响力的用户，适用于有向和无向网络。LKG 混合方法包括三个阶段：（1）社区检测，其中使用Louvain算法将完整的社交网络划分为相关社区 ；(2) 通过在每个划分的社区局部应用k-shell 分解来检测社区顶级节点 ；(3) 选择泛化，其中先前获得的结果是泛化以最大限度地扩大影响力。对几个不同大小的数据集进行了实验研究。结果表明，与现有的相关工作模型相比，使用不完整的社交网络在影响力的传播方面取得了更好的结果，并且处理时间要少得多。