Research on motif discovery algorithm in network based on MapReduce

2nd International Symposium on Application of Materials Science and Energy Materials

材料科学;能源学

Liu, Zheng^1 ; Zhang, Qian^2

College of Computer Science and Engineering, Northeastern University, China^1

Advanced Product Division, Neusoft Corporation, China^2

关键词: Expandability;    In networks;    Large-scale network;    Motif discovery;    Parallel programming model;    Serial algorithms;    Single- machines;    Software network;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/490/4/042026/pdf
DOI  :  10.1088/1757-899X/490/4/042026

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

The motif discovery algorithms in network are always serial algorithms that run on single machine, which leads to lower efficiency and cannot meet the demands for discovering motifs in large-scale networks. In response to this situation, a parallel motif discovery algorithm is presented based on a parallel programming model called MapReduce in this paper. The algorithm is realized on Hadoop to discover the motifs in software networks with different scale. From the application test, the efficiency, speedup and expandability of the parallel algorithm are analysed and verified.