An Algorithm for Detection of Shared Communities in Social Networks

Abstract

Social networks are made up of a variety of actors and their interactions. Humans, blogs, newlinephotographs, Instagram photographs, news stories, and websites may all be considered as newlineactors. Human communication is growing increasingly reliant on social media. For the newlinepast several years, numerous academics and scientists have been interested in the structure newlineand evolution of various real-world networks, as well as their study. newlineA network s community may be described as a group of nodes that are more tightly newlineconnected to one another than the rest of the network s nodes. Clustering (or community newlinestructure) is a phenomena that may be seen not just in naturally occurring networks, newlinebut also in man-made networks such as the World Wide Web. The goal of community newlinedetection is to discover groupings of nodes from a graphical structure that represents a newlinefunctional unit in some way. Community detection is useful for the applications, such newlineas, detecting suspicious events in telecommunication networks, refactoring the software newlinepackages, recommendation systems for online-shopping and entertainment, link prediction, newlinecontrolling epidemic spreading, detecting terrorist and criminal groups, and controlling newlineinformation diffusion. However, community detection is a difficult task because there newlineare numerous definitions for community based on various parameters; finding an optimal newlinecommunity based on a scoring function is an NP-hard problem; communities in social newlinenetworks tend to overlap to varying degrees; and selecting suitable parameters to evaluate newlinedetected communities in the absence of ground truth data is a difficult task. newlineFor real-world networks, the early random graph models for social networks were newlineproven to be inefficient. Preferential attachment and community structure models are the newlinemost realistic network models. Because of the preferred attachment paradigm, certain newlinenodes operate as hubs, sharing the majority of the linkages. The preferred attachment newlineconcept is to responsible for the social network s shrinking width and av