Taking advantage of the structure of logical representations, in this paper we report an algorithm that evaluates conjunctive queries in a massively parallel environment under an object-based representation for deductive databases. By distributing objects in a database, we show that parallel evaluation of a query can be achieved in a cooperative way so that the conventional tuple-by-tuple, operation-by-operation evaluation strategy can be replaced by a global, parallel matching approach. With the proposed scheme, all conjuncts of a given query can be examined at the same time, which enables us to eliminate the need of any temporary relation. On the other hand, compared with the interpretive method, we show that any data dependency imposed by shared variables is no longer a major problem in achieving AND-parallelism by the proposed scheme.
|Number of pages
|IEEE Transactions on Knowledge and Data Engineering
|Published - 1994 Feb
All Science Journal Classification (ASJC) codes
- Information Systems
- Computer Science Applications
- Computational Theory and Mathematics