New metaheuristic algorithms for the analysis of the user influence in social networks

Abstract

The evolution and spread of social networks have attracted the interest of the scientific community in the last few years. Specifically, several new interesting problems which are hard to solve have arisen in the context of viral marketing, disease analysis and influence analysis , among others. Companies and researchers try to find the elements that maximize profit, stop pandemics, etc. This family of problems are usually known as Social Network Influence Maximization (SNIM) problems , whose goal is to find the most influential users (commonly known as seeds) in the social network, simulating an influence diffusion model. Since SNIM is known to be an NP-hard problem, and most of the networks to analyze are considerably large, this problem have become a real challenge for the scientific community. Different works have attempted to solve it by means of heuristic or metaheuristic approaches, such as evolutionary algorithms based on simulations of the spreading mechanism , but they are not able to solve real-world large-scale networks due to their limitations in computing time. The main drawback of this optimization problem lies in the computational effort required to evaluate a solution. Since the infection of a node is evaluated in a probabilistic manner, the objective function value requires from a Monte Carlo simulation to see the robustness of the method, resulting in a computationally complex process. The current proposal tries to overcome this limitations by considering a metaheuristic algorithm based on the Greedy Random-ized Adaptive Search Procedure (GRASP) framework. The construction phase is based on static features of the network, which notably increases its efficiency, since it does not require to perform any simulation during construction. The local improvement involves a surrogate-assisted-swap local search to find the most influential users based on swap moves. Experiments performed on 6 well-known social networks datasets with 5 different seed set sizes confirm that the proposed algorithm is able to provide competitive results in terms of quality and computing time when comparing it with the best algorithms found in the state of the art.

Publication
16º International Conference on Parallel Problem Solving from Nature (PPSN), Workshop on Evolutionary and Bio-inspired techniques for Social Network Analysis (BioSocNets)
Isaac Lozano-Osorio
Isaac Lozano-Osorio
Artificial Intelligence Phd Student

Isaac Lozano graduated with a double degree in Computer Engineering and Computer Engineering from the Universidad Rey Juan Carlos, where he was awarded the prize for the Best Final Project. Subsequently, he completed a Master in Artificial Intelligence Research (UIMP). His main research interests are focused on the interface between Computer Science, Artificial Intelligence and Operations Research. Most of his publications deal with the development of metaheuristic procedures for graph modeled optimization problems.

Jesús Sánchez-Oro
Jesús Sánchez-Oro
Associate Professor

Associate Professor at the Computer Science Department, being one of the senior researchers of the Group for Research on Algorithms For Optimization GRAFO.

Abraham Duarte
Abraham Duarte
Full Professor

Abraham Duarte is Full Professor in the Computer Science Department at the Rey Juan Carlos University (Madrid, Spain). He has done extensive research in the interface between computer science, artificial intelligence, and operations research to develop solution methods based on Computational Intelligence (metaheuristics) for practical problems in operations-management areas such as logistics and supply chains, telecommunications, decision-making under uncertainty and optimization of simulated systems.