The Influence of Location on Nodes’ Centrality in Location-Based Social Networks

Nowadays, due to the widespread use of social networks, they can be used as a convenient, low-cost, and affordable tool for disseminating all kinds of information and data among the massive users of these networks. Issues such as marketing for new products, informing the public in critical situations, and disseminating medical and technological innovations are topics that have been considered by the owners of companies as well as government and private organizations. In order to achieve this goal, the methods of maximizing the speed and quality of information dissemination in social networks have been studied by researchers in various fields of science. One of the best ways to increase the speed of information diffusion is to take advantage of the social status of influential node. The role of some people within the networks is more prominent than the others, who are known by titles such as leader, important, influential, central, and vital. These people having the centrality in the social networks can have the greatest impact on the network. Identifying these people allows us to control the way of information dissemination, the spread of diseases, the more effective advertising of commercial products, or the identification of the head of different social groups. Different measures have been proposed to identify these people and to measure the centrality of nodes in a network. Identifying influential people in a network is not easy task, and the criteria for being important are varied. Therefore, there is no general index that best determines the importance of people in a network, and this index varies from one network to another and from one situation to another. In addition, whether these indexes are local or global can be a point of contention and it can produce different concepts of centrality. This paper aims to provide a new measure so-called “socio-spatial centrality” to find the central nodes in the location-based social networks, which is a linear combination of social centrality and spatial centrality. In the new proposed measure, the central node is the one that, while having more direct neighbors than the other nodes, its neighboring nodes will have a uniform and proper spatial dispersion within the given area. Degree centrality has been used to determine the social centrality and the standard distance is used to calculate the spatial centrality. By simultaneously assessing the nodes' degree and the spatial dispersion of their neighbors, the top-k nodes are identified as influential nodes. The proposed measure, along with the existing methods for measuring centralities such as degree centrality, betweenness centrality, and closeness centrality, has been applied to two real location-based social networks. The results show that the geographic coverage of the selected nodes by the socio-spatial centrality was the highest in comparison to other centrality measures, and the correlation between the output of the new centrality measure as well as the conventional centrality measures is decreasing with the increase in the number of selected nodes. Overall, evaluations show the positive impact of the proposed method on the geographic coverage maximization in location-based social networks.