跳到主要导航 跳到搜索 跳到主要内容

A feasible graph partition framework for random walks implemented by parallel computing in big graph

  • Xi'an Jiaotong University
  • Tsinghua University

科研成果: 书/报告/会议事项章节会议稿件同行评审

3 引用 (Scopus)

摘要

Graph partition is a fundamental problem of parallel computing for big graph data. Many graph partition algorithms have been proposed to solve the problem in various applications, such as matrix computations and PageRank, etc., but none has pay attention to random walks. Random walks is a widely used method to explore graph structure in lots of fields. The challenges of graph partition for random walks include the large number of times of communication between partitions, too many replications of the vertices, unbalanced partition, etc. In this paper, we propose a feasible graph partition framework for random walks implemented by parallel computing in big graph. The framework is based on two optimization functions to reduce the bandwidth, memory and storage cost in the condition that the load balance is guaranteed. In this framework, several greedy graph partition algorithms are proposed. We also use five metrics from different perspectives to evaluate the performance of these algorithms. By running the algorithms on the big graph data set of real world, the experimental results show that these algorithms in the framework are capable of solving the problem of graph partition for random walks for different needs, e.g. the best result is improved more than 70 times in reducing the times of communication.

源语言英语
主期刊名Proceedings of the 34th Chinese Control Conference, CCC 2015
编辑Qianchuan Zhao, Shirong Liu
出版商IEEE Computer Society
4986-4991
页数6
ISBN(电子版)9789881563897
DOI
出版状态已出版 - 11 9月 2015
活动34th Chinese Control Conference, CCC 2015 - Hangzhou, 中国
期限: 28 7月 201530 7月 2015

出版系列

姓名Chinese Control Conference, CCC
2015-September
ISSN(印刷版)1934-1768
ISSN(电子版)2161-2927

会议

会议34th Chinese Control Conference, CCC 2015
国家/地区中国
Hangzhou
时期28/07/1530/07/15

学术指纹

探究 'A feasible graph partition framework for random walks implemented by parallel computing in big graph' 的科研主题。它们共同构成独一无二的指纹。

引用此