欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
社会网络中影响最大化问题是指在特定传播模型下,获取一个指定大小的节点集合,使得该集合在网络中的聚合影响力最大.针对贪心算法运用于大规模社会网络时存在效率低下且不可扩展的问题,文中提出基于核数层次特征和影响半径的启发式算法——核覆盖算法（Core Covering Algorithm,CCA）.该算法首先引入k-核概念,基于k-核分解求出每个节点的核数,然后根据核数分布的层次性,引入节点的影响半径参数,最后综合核数和度数两个属性,找出影响力节点集合.文中在两个数据集和两种传播模型上进行了实验,结果表明：（1）在传播概率较大的独立级联模型（Independent Cascade Model,IC）下,CCA能取得比现有启发式算法更优的影响效果;（2）在三价（TRIVALENCY Model,TR）模型下,CCA的表现也同样优于其他启发式算法;（3）与其他启发式算法相比,CCA的运行时间更少.
英文摘要：
Influence maximization is the problem of obtaining a set of nodes with specified size in a social network to maximize their aggregate influence under certain influence diffusion model. Since greedy algorithms are inefficient and not-scalable, we propose a heuristic algorithm named Core Covering Algorithm （CCA） based on the coreness hierarchical characteristic and influence radius. Firstly, the algorithm introduces the concept of k-core and calculates the coreness of each node. Then, it introduces the influence radius parameter according to the hierarchy of coreness. Finally, it identifies influential nodes in accordance with the coreness and degree. Experiments are conducted on two datasets and two diffusion models. Experimental results show that （1） CCA performs better than other heuristic algorithms under Independent Cascade Model with a larger influence probability; （2） CCA also performs better than other heuristic algorithms under TRIVA- LENCY model; （3） Compared with other heuristic algorithms, CCA has lower running time.
同期刊论文项目
同项目期刊论文
Towards optimized scheduling for data-intensive scientific workflow in multiple datacenter environme
Scientific workflow scheduling in non-dedicated heterogeneous multicluster with advance reservations
期刊信息
位置：