Efficient Matching and Merging for Top-K Graph Pattern Mining on the Cloud

• Main Authors: Kuan-Wei Lee, 李冠緯
• Other Authors: Ming-Syan Cheng
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/444es6

碩士 === 國立臺灣大學 === 電機工程學研究所 === 101 === Mining large structural patterns in graph data is an important problem in data mining research area. It has been applied applied in many domains such as social media, bioinformatics, and chemical drugs. Due to the rapidly increasing large scale graph data sets in recent years, traditional algorithms cannot process the big graph data. Although many distributed graph pattern mining algorithms have been proposed to solve the big data problem, the existing distributed algorithms still suffer from many problems on distributed pattern mining, such as expensive sharing information cost among slaves, insuffcient scalability and, lower accuracy. To overcome these problems, we propose a distributed graph pattern mining algorithm for mining top-k large structural patterns in the cloud computing environment. We propose the partitioning and merging methodology to improve the scalability and the efficiency.