United Kingdom and

The importance of query optimisation has led to active research in this area over a number of years and the development of optimisation components as a standard feature of large Database Management Systems. More recent work is concerned with semantic query optimisation (SQO) which is guided by derived knowledge (rules), about the data itself, using techniques associated with data mining. Fast, efficient rule generation and maintenance is crucial to achieving high optimiser performance. At the same time it is recognized that most available computing power, in a networked client-server system, resides in the client workstations which are often under-utilized and that this resource is expanding rapidly with each successive new generation of workstation. In this paper we show how this spare capacity may be used to improve optimiser performance by distributing the rule management processes in a parallel virtual machine environment.