Splashdot Case: A New Pagerank Algorithm For Influence Maximization In Signed Social Networks

Abstract

Understanding and maximizing the information’s dynamics flow in social networks requires consideration of influence maximization. Traditional influence maximizing techniques, on the other hand, tend to emphasize good associations and downplay the significance of negative ones, which limits their application in real-world situations where friend or foe interactions coexist. This thesis provides "Signed-PageRank," a fresh and effective framework made to solve impact maximization in signed social networks, as a solution to this issue. Utilizing the idea of signed edges, the Signed-PageRank architecture take into account both favorable and unfavorable engagements between nodes. This thesis provides a modified PageRank method that considers the impact of signed edges in order to precisely represent the dynamics of influence inside the network. Our method offers a more thorough understanding of patterns of influence propagation and the creation of attitudes in complex social networks by including both the positive and negative interactions. In addition, this thesis provides a scalable technique to quickly calculate the Signed PageRank scores, allowing us to recognize significant nodes and their related impact values in huge signed social networks. This thesis undertakes comprehensive tests on real-world dataset of Slashdot to show the efficacy of our approach. As a result, the Signed-PageRank architecture we've described represents a major improvement in impact maximizing for social networks with signed edges. It brings us new perspectives for comprehending information spread, sentiment propagation, and behavioral dynamics in actual social systems by taking into account interactions involving both favorable and unfavorable elements.