Towards Scalable Graph Analytics on Time Dependent Graphs
Towards Scalable Graph Analytics on Time Dependent Graphs
复制标题
面向时间相关图的可扩展图分析
DOI：
--
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
M. Gokhale
中科院分区：
文献类型：
--
作者：
S. Poudel;R. Pearce;M. Gokhale
时间相关图中的边参数随时间变化。在这种情况下的图形分析考虑了图形的实际视图随时间变化的事实。作为示例，时间相关图中的单源最短路径（SSSP）分析随着图遍历的不同开始时间和/或遍历期间每个节点处的不同等待时间的最佳到达时间。时间相关图的分析大多按顺序应用于各种应用中[1, 2]。在这项工作中，我们评估了 HPC 系统上大型分布式图的图分析。
Edge parameters in time dependent graphs vary as a function of time. Graph analytics in such context takes into consideration the fact that the actual view of the graph changes with time. As an example, single source shortest path (SSSP) in time dependent graphs analyses optimal arrival time with varying start time of graph traversal and/or varying waiting time at each node during traversal. Analytics in time dependent graphs have been applied mostly sequentially in various applications[1, 2]. In this work, we evaluate the graph analytics in large distributed graphs on HPC systems.