Adapting Eclat algorithm to parallel environments with Charm++ library

Marek Puścian , Waldemar Grabski

Abstract

In this paper we describe Eclat algorithm that is adapted to deal with growing data repositories. The presented solution utilizes Master-Slave scheme to distribute data mining tasks among available computation nodes. Several improvements have been proposed and successfully implemented using Charm++ library. This paper introduces optimization techniques to reduce communication cost and synchronization overhead. It also discusses results of the performance of parallel Eclat algorithm against different databases and compares it with parallel Apriori algorithm. The proposed approach has been illustrated with many experiments and measurements performed using multiprocessor and multithreaded computer platform.
Author Marek Puścian (FEIT / IN)
Marek Puścian ,,
- The Institute of Computer Science
, Waldemar Grabski (FEIT / IN)
Waldemar Grabski,,
- The Institute of Computer Science
Pages100315A-1-100315A-8
Publication size in sheets0.5
Book Romaniuk Ryszard (eds.): Proc. SPIE. 10031, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2016, vol. 10031, 2016, SPIE , ISBN 9781510604858, [781510604865 (electronic) ], 1170 p., DOI:10.1117/12.2257157
Keywords in Englishalgorithm parallelization, Eclat, Charm++, frequent pattern mining, speed up measurement
DOIDOI:10.1117/12.2249403
URL http://dx.doi.org/10.1117/12.2249403
ProjectDevelopment of new algorithms in the areas of software and computer architecture, artificial intelligence and information systems and computer graphics . Project leader: Rybiński Henryk, , Phone: +48 22 234 7731, start date 18-05-2015, end date 30-11-2016, II/2015/DS/1, Completed
WEiTI Działalność statutowa
Languageen angielski
File
100315A.pdf 294.13 KB
Score (nominal)15
ScoreMinisterial score = 15.0, 25-09-2019, BookChapterMatConfByConferenceseries
Ministerial score (2013-2016) = 15.0, 25-09-2019, BookChapterMatConfByConferenceseries
Citation count*
Cite
Share Share

Get link to the record


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back
Confirmation
Are you sure?