Centrality measure in social networks based on linear threshold model

Abstract

Centrality and influence spread are two of the most studied concepts in social network analysis. In recent years, centrality measures have attracted the attention of many researchers, generating a large and varied number of new studies about social network analysis and its applications. However, as far as we know, traditional models of influence spread have not yet been exhaustively used to define centrality measures according to the influence criteria. Most of the considered work in this topic is based on the independent cascade model. In this paper we explore the possibilities of the linear threshold model for the definition of centrality measures to be used on weighted and labeled social networks. We propose a new centrality measure to rank the users of the network, the Linear Threshold Rank (LTR), and a centralization measure to determine to what extent the entire network has a centralized structure, the Linear Threshold Centralization (LTC). We appraise the viability of the approach through several case studies. We consider four different social networks to compare our new measures with two centrality measures based on relevance criteria and another centrality measure based on the independent cascade model. Our results show that our measures are useful for ranking actors and networks in a distinguishable way.Peer ReviewedPostprint (author's final draft

Critical Impact of Social Networks Infodemic on Defeating Coronavirus COVID-19 Pandemic: Twitter-Based Study and Research Directions

“…Accordingly, extensive effort should be put to elaborate methodologies and recommendation systems for efficiently recognizing credible and convincing influencers in specific events/locations/communities (e.g., based on profile, insights, historical engagement) for spreading the relevant and cited information provided by trusted scientists and experts at large scale, in the right place and to the right people. In this context, researchers may benefit from the rich literature that targets identifying influencers based on selected events in order to build relevant approaches [47]- [54]. • Current raised infodemic shed the light on the urgent need to elaborate methodologies and techniques to be embedded in the social network platforms for systematically adopting emergency and crisis mode management strategies and responding to the situation dangers.…”

“…Such techniques stem from the need to rank the relevance of the users and their tweets, and thus two main categories of solutions exist to address the issue in dispute. The first set of approaches focused on the content to assign reputation using machine learning techniques [47], [58], [59], while the second set relied on the user and its relation described as nodes in a graph model [35], [48], [58]. Moreover, there are other solutions that depend on both methods to achieve better accuracy.…”

Identifying Opinion Leaders on Social Networks Through Milestones Definition

“…Regarding the usefulness of the centrality measures, there are two important factors that have been considered in other studies to determine if the users are ranked properly, namely, the standard deviation (σ ) and the number of different results obtained by the measure [13]. On one hand, a large number of different results implies that the measure generates distinguishable classes of users with different values.…”

“…To address this problem, numerous centrality measures have been defined, providing different criteria to automatically rank the users of the network [9]. Some influence measures are based on simple metrics related to user activity within the network; some others are based on the PageRank algorithm [10], which is in turn inspired on the eigenvector centrality [11]; some others in the influence spread phenomenon [12], [13]; and others in predictive algorithms [9].…”

Containment of rumor spread in complex social networks

“…The authors then presented a greedy algorithm to solve the problem that can achieve a good sub-optimal solution. Following this work, improved greedy algorithms [8,9,13,24], community based approaches [7,23,36], new centrality measures [35,37] and efficient heuristics [15,25,31,45] have been proposed to solve the problem aiming to balance the time complexity of algorithms and the influence propagation and trying to make them scalable to large datasets. In recent years, researchers have studied the influence maximization problem in more complex networks by taking the heterogeneity of individual relationships [18] or multiplexity [42] into consideration.…”