Parallel programming: techniques and applications using networked workstations and parallel computers
MPI: The Complete Reference
Parallel DSIR Text Retrieval System
Proceedings of the 6th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
The text retrieval conferences (TRECS)
TIPSTER '98 Proceedings of a workshop on held at Baltimore, Maryland: October 13-15, 1998
Hi-index | 0.00 |
In this paper, we present another study of an improved parallel DSIR text retrieval system that can perform fast indexing of several gigabytes of text collection using Pentium-class PC-cluster. We use multiple-master/slave principle to implement this parallel indexing algorithm. In a computing node, a master process has been designed to work in conjunction with each slave process in order to utilize as much as possible the computing power during one or another process is waiting for I/O. We also present special buffering and caching techniques for boosting the computing performance. We have tested this algorithm and presented the experimental results using a large-scale TREC8 collection and investigated both computing performance and problem size scalability issue.