Design and Analysis of Different Perspectives for Signed Social Networks Using Nature Inspired Techniques

Abstract

Social networks shows interpersonal connections between different people, such as friendships and shared interests. Social network analysis examines these social networks. relationships. Algorithms for link prediction are used to forecast these interpersonal connections. Presented with a social network graph, in which a user is represented by a node, and the user relationships, a link prediction method, forecasts the potential new connections that may be made in the upcoming. Social networks are extensive systems that show the connections between countless social elements. One of the main research areas of social network analysis and network analysis is the study of patterns and evolution. A component of this problem is the link prediction problem, which is a way to predict future associations between unconnected nodes. Traditional approaches are made to operate with social networks in a certain context. However, the data from these networks is frequently erratic, absent, and prone to observation errors that lead to deformations and probably unreliable results. The belief function theory, a compelling paradigm for reasoning under uncertainty that allows for the representation, quantification, and management of faulty information, is used in this research to address the link prediction problem. First, a brand- new graph based social network model that takes into account link structural uncertainty is presented. The belief functions tools are then used to present a novel approach for the prediction of new relationships. In order to forecast new connections, it makes use of neighbourhood and shared group information in social networks. The effectiveness of the new method was tested on real social networks. Studies have shown that our strategy outperforms existing methods based on structural information.