图数据分析系统计算模型综述

Survey on computing models in graph data analysis systems

为适应图数据规模巨大、耦合性强、动态变化等特点,实现大规模图数据的高效分析计算,对图计算系统计算模型的研究现状进行了调研和综述。介绍了图计算系统的产生和发展,然后将主流图计算系统中的计算模型按照计算对象分为节点中心计算模型、边中心计算模型、路径中心计算模型和子图计算模型四类,重点介绍节点中心模型的应用和性能。最后对图计算模型的发展过程进行总结,并展望图计算模型未来的发展方向。

In order to adapt to the characteristics of graph data, included large scale, strong coupling, dynamic change,and so on, and realize the efficient analysis and calculation of large scale graph data, this paper investigated the current research Status of computing model in graph computing systems. Firstly,it introduced the development of graph computing systems. And then it classified computing models in major graph computing systems into four categories, i. e. vertex-centric computing model, edgecentric computing model, path-centrie computing model and subgraph-centric computing model, and introduced the application and performances of vertex-central computing models. Finally, it summarized the development of graph computing model, and prospected the future research direction.