December 30, 2002
View Full PaperPerformance of inverted indices in shared-nothing distributed text document information retrieval systems
Key Points
Key points are not available for this paper at this time.
Abstract
The impact on query processing performance of various physical organizations for inverted lists is compared. A probabilistic mode of the database and queries is introduced. Simulation experiments determine which variables most strongly influence response time and throughput. This leads to a set of design tradeoffs over a range of hardware configurations and new parallel query processing strategies.>