Optimizing Chain Queries in a Distributed Database System

Publisher Website
Google Scholar
Abstract
[[abstract]]The problem of optimal query processing in distributed database systems was shown to be NP-hard. However, for special types of queries called star queries, we have developed a polynomial optimal algorithm. In an earlier paper, we described an approach to obtain the optimal semi-join program for a star query by gradually reducing the search space to a minimal set S without making any assumptions on the file sizes and the semi-join selectivities. In this paper, by making certain assumptions on the file sizes and the semi-join selectivities, the size of S can be reduced to unity; i. e. , given a star query, we can directly generate the optimal program. We include an example which compares the performance of existing heuristic algorithms with our proposed optimal algorithm.[[fileno]]2030208030105[[department]]資訊工程學

This publication has 5 references indexed in Scilit: