Mineração de dados paralela e distribuída baseada no ambiente Weka

Abstract

Technological development have allowed the generation and recording of data volumes increasing. Data mining consists in apply algorithms over large data bases to extract them useful knowledge. The database size and the complexity of the techniques involved makes necessary use of distributed solutions in the data mining process. WEKA (Waikato Environment for Knowledge Analysis) is a centralized data mining environment that has been used as a basis to parallel and distributed data mining tools. The purpose of this work is explore, through a case study, Grid WEKA, a tool based on WEKA. In this context are identified included techniques, the available of parallelism and distribution and analyzed and discussed the performance using different data mining techniques in different aspects of the distributed environment.