Influence Maximization in social networks is defined as determining a subset of seed nodes where triggering the Influence diffusion through the social network leads to the maximum number of final Influenced nodes. The tradeoff between the runtime efficiency and effectiveness in the quality of response is the main issue in presenting solutions for this NP-hard optimization problem. Centrality-based methods are applied as a category of efficient heuristic-based solutions to this problem. The two leading causes of losing effectiveness in centrality-based methods are 1) only the link structure and non-awareness of Influence diffusion are considered in determining the importance of nodes, and 2) Influence overlap exists among selected seed nodes. To address the first cause, an Influence-aware betweenness centrality measure is proposed considering both IC and LT models. Moreover, an existing Influence-aware closeness centrality measure for LT model is adopted to cover both LT and IC models. To address the second cause, a greedy-based method is proposed by applying Influence-aware centrality measures to identify the influential nodes, next to proposing a Jacquard-based measure to overcome the Influence overlap problem. The experiments consist of two parts where two real-world datasets are applied: 1) the proposed Influence-aware centrality measures are compared with their original versions, and 2) the greedy-based method is compared with benchmark methods. The results indicate the effectiveness of the Influence-aware centrality measures and the proposed greedy-based method in maximizing the Influence spread in social networks.

Many real-world networks, including biological networks, internet, information and social networks can be modeled by a complex network consisting of a large number of elements connected to each other. One of the important issues in complex networks is the evaluation of node importance because of its wide usage and great theoretical significance, such as in information diffusion, control of disease spreading, viral marketing and rumor dynamics. A fundamental issue is to identify a set of most influential individuals who would maximize the Influence spread of the network. In this paper, we propose a novel algorithm for identifying influential nodes in complex networks with community structure without having to determine the number of seed nodes based on genetic algorithm. The proposed algorithm can identify influential nodes with three methods at each stage (degree centrality, random and structural hole) in each community and measure the spread of Influence again at each stage. This process continues until the end of the genetic algorithm, and at the last stage, the most influential nodes are identified with maximum diffusion in each community. Our community-based Influencers detection approach enables us to find more influential nodes than those suggested by page-rank and other centrality measures. Furthermore, the proposed algorithm does not require determining the number of k initial active nodes.

The Influence Maximization problem in social networks aims to find a minimal set of individuals in order to produce the highest Influence on the other individuals in the network. In the last two decades, a lot of algorithms have been proposed to solve the time efficiency and effectiveness challenges of this NP-Hard problem. Undoubtedly, the CELF algorithm (besides the naive greedy algorithm) has the highest effectiveness among them. Of course, the CELF algorithm is faster than the naive greedy algorithm (about 700 times). This superiority has led many researchers to make extensive use of the CELF algorithm in their innovative approaches. However, the main drawback of the CELF algorithm is the very long running time of its first iteration since it has to estimate the Influence spread for all nodes by the expensive Monte-Carlo simulations, similar to the naive greedy algorithm. In this paper, a heuristic approach is proposed, namely optimized-CELF algorithm, in order to improve this drawback of the CELF algorithm by avoiding the unnecessary Monte-Carlo simulations. It is found that the proposed algorithm reduces the CELF running time, and subsequently, improves the time efficiency of the other algorithms that have employed CELF as a base algorithm. The experimental results on the wide spectral of real datasets show that the optimized-CELF algorithm provides a better running time gain, about 88-99% and 56-98% for k=1 and k=50, respectively, compared to the CELF algorithm without missing effectiveness.

During the very last decade, people have been spending lots of time working with social networks to interact with friends and to share information, thoughts, news, and etc. These social networks comprise a very important part of our daily lives. Along with the exploitation of the development of social networks, finding influential individuals in a social network has many practical functions in marketing, politics, and even control of the diseases. In the present research, a novel method called the dynamic generalized vulture algorithm has been proposed to solve Influence Maximization problems. Regarding the fact that in real world social networks own very dynamic and scalable nature, through our proposed algorithm, we have considered two important criteria which have been rarely taken into consideration in previous projects. The first criterion is due to the network structure change during time pass and the other refers to scalability. The suggested algorithm was measured considering standard data sets. The results showed that the proposed algorithm has been more scalable and has had higher precision in locating the most influential tops in such networks compared with other algorithms due to the reduction of search area and using several different mechanisms during navigation and optimization, balance creation and moving through these stages.

Influence Maximization serves as the main goal of a variety of social network activities such as viral marketing. The independent cascade model for the Influence spread assumes a one-time chance for each activated node to Influence its neighbors. On the other hand, the manually activated seed set nodes can be reselected without violating the model parameters or assumptions. This view divides the Influence Maximization process into two cases: the simple case where the reselection of the nodes is not considered and the reselection case. In this study we will analyze real world networks in the reselection case. First we will show that the difference between the simple and the reselection cases constitutes a wide spectrum of networks ranging from the reselection-free to the reselection-friendly ones. Then we will experimentally show a significant entanglement between this and Influence spread dynamics as well as other structural parameters of the network. Specifically, we show that under a realistic condition, the reselection gain of a network has a correlation of 0. 73 to a newly introduced Influence spread dynamic. Furthermore, we propose a measure for detecting star-like networks and experimentally show a significant correlation between our proposed measure and the reselection gain in real world networks with different edge weight models.