{"title":"A Multiobjective Evolutionary Approach for Influence Maximization in Multilayer Networks","authors":"Qipeng Lu, Zhan Bu, Yuyao Wang","doi":"10.1145/3404555.3404568","DOIUrl":null,"url":null,"abstract":"Influence Maximization (IM) is one key algorithmic problems in information diffusion research; it aims to select a set of users from a social network and, by following a specific model, maximize the number of users influenced (the influence spread). Yet despite its immense potential, relatively little research is dedicated to IM for multilayer networks. Conversely, most existing IM studies that rely on a greedy algorithm strategy only obtain a single solution that provides limited insights on the target networks' core organization. With that in mind, we focus on studying the Influence Maximization Problem (IMP) in multilayer networks. Specifically, we define novel concepts, such as the pairwise reciprocal length and pairwise influence, with respect to the information-diffusion process in multilayer networks. Then we formulate the IM in multilayer networks as a multiobjective optimization problem and employ the classic Nondominated Sorting Genetic Algorithm II (NSGA-II) to find a set of Pareto-optimal solutions that provide a wide range of options for decision makers. To maintain population diversity and accelerate the algorithm's convergence, we combine a heuristic population initialization strategy and an efficient two-point crossover operation. Extensive experiments show that our approach has competitive performance when compared to off-the-shelf IM algorithms with regard to influence spread and running time.","PeriodicalId":220526,"journal":{"name":"Proceedings of the 2020 6th International Conference on Computing and Artificial Intelligence","volume":"47 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 2020 6th International Conference on Computing and Artificial Intelligence","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3404555.3404568","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
Influence Maximization (IM) is one key algorithmic problems in information diffusion research; it aims to select a set of users from a social network and, by following a specific model, maximize the number of users influenced (the influence spread). Yet despite its immense potential, relatively little research is dedicated to IM for multilayer networks. Conversely, most existing IM studies that rely on a greedy algorithm strategy only obtain a single solution that provides limited insights on the target networks' core organization. With that in mind, we focus on studying the Influence Maximization Problem (IMP) in multilayer networks. Specifically, we define novel concepts, such as the pairwise reciprocal length and pairwise influence, with respect to the information-diffusion process in multilayer networks. Then we formulate the IM in multilayer networks as a multiobjective optimization problem and employ the classic Nondominated Sorting Genetic Algorithm II (NSGA-II) to find a set of Pareto-optimal solutions that provide a wide range of options for decision makers. To maintain population diversity and accelerate the algorithm's convergence, we combine a heuristic population initialization strategy and an efficient two-point crossover operation. Extensive experiments show that our approach has competitive performance when compared to off-the-shelf IM algorithms with regard to influence spread and running time.