The Preliminary Design of Papyrus: A System for High Performance, Distributed Data Mining over Clusters, Meta-Clusters and Super-Clusters (1999)
| Venue: | In Proceedings of Workshop on Distributed Data Mining, alongwith KDD98 |
| Citations: | 21 - 5 self |
BibTeX
@INPROCEEDINGS{Grossman99thepreliminary,
author = {Robert Grossman and Stuart Bailey and Ashok Ramu and Balinder Malhi and Andrei Turinsky},
title = {The Preliminary Design of Papyrus: A System for High Performance, Distributed Data Mining over Clusters, Meta-Clusters and Super-Clusters},
booktitle = {In Proceedings of Workshop on Distributed Data Mining, alongwith KDD98},
year = {1999},
pages = {259--275}
}
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
Data mining is a problem for which cluster computing provides a competitive alternative to specialized high performance computers for mining large data sets. Distribued clusters provide a natural infrastructure for mining large distributed data sets. Distributed clusters can be connected by commodity networks to form what we call meta-clusters and by high performance networks to form what we call super-clusters. In this paper, we describe the design of a system called Papyrus which is designed for mining data which is distributed over clusters, meta-clusters, and super-clusters. We also describe some experimental results of a preliminary implementation.
