Skip to Main Content
This paper presents a development of DWORM, a prototype of a distributed information retrieval system based on the Grid environment. The Grid environment enables traditional information retrieval systems to manage rapidly a voluminous amount of geographically distributed Web data by massively computing collaboration and resource sharing. Using a divide-and-conquer strategy, the document indexing and retrieval processes are disseminated and performed on each computing node. Preliminary evaluation done on machines under the ApGrid shows that DWORM scales optimally with four worker nodes; 45 percent more performance is obtained than the sequential WORM.