2009 Fiscal Year Final Research Report

Research on Parallel Extraction of Frequent Sequence Pattern on a Multi-Core Computer Cluster.

Research Project

Project/Area Number	20700095
Research Category	Grant-in-Aid for Young Scientists (B)
Allocation Type	Single-year Grants
Research Field	Media informatics/Database
Research Institution	Hiroshima City University
Principal Investigator	TAMURA Keiichi Hiroshima City University, 情報科学研究科, 講師 (80347616)
Project Period (FY)	2008 – 2009
Keywords	並列処理 / マルチコア / データマイニング
Research Abstract	This study has developed a novel parallelization model for extraction of frequent sequence pattern on a computer cluster which is constructed by computers with multi-core processors, which is becoming mainstream of CPU. In order to use computer resources efficiently on a multi-core computing cluster, the developed parallelization model has three characteristics, (1) efficient use method of shared resource, (2) efficient coordinated technique between cores and between computers, (3) dynamic load-balancing technique for a multi-core computer cluster. A typical frequent pattern extraction algorithm was made parallel by using this model. The experimental results shows good performance.

(3 results)