基于层次化社区结构的影响最大化算法  

Hierarchical community structure based algorithm for influence maximization

作　　者：朱昌盛[1] 朱福喜[1] 阳小兰[2] 

机构地区：[1]武汉大学计算机学院,湖北武汉430072 [2]武昌理工学院信息工程学院,湖北武汉430223

出　　处：《计算机工程与设计》2017年第3期561-565,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(61272277);湖北省自然科学基金项目(2014CFB356)

摘　　要：针对贪心算法时间复杂度过高,不适用于大型社会网络问题,提出基于层次化社区结构的影响最大化算法HCSA(hierarchical community structure algorithm)。对社会网络进行层次化社区划分,启发式地选择覆盖率最大的社区层级并按比例分配初始节点,综合节点度数,得出初始节点集合。实验结果表明,与现有启发式算法相比,HCSA算法取得了更广的传播范围和更少的运行时间。Since the time complexity of the greedy algorithm is too high,and the algorithm does not apply to large social network,hierarchical community structure based algorithm（HCSA）was proposed for influence maximization.The social network was divided into hierarchical communities,community-level with the largest coverage rate was chosen and initial nodes were allocated heuristically.The degree was considered and initial nodes set was obtained.Experimental results show that wider scope of spread and less running time were gained using HCSA.

关 键 词：社会网络 层次化 社区结构 影响最大化 独立级联模型 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]