Distributed query optimization using reduction filters.

The optimization of general queries in a distributed database management system is an important research topic. The difficulty is to select the database operations which will process the query and minimize costs. Traditional solutions include the use of heuristic strategies based on semijoin or join operations. Here, we present an approach for general queries which uses reduction filters, which are based on Bloom filters, to minimize data transfers and reduce local processing costs. We discuss related work and present our algorithm - illustrating it with a simple example. The paper ends with a brief discussion of current and future work.