|
Abstract : |
Managing data in large rule systems is a critical issue, and DBMS systems are being extended for the support of rule-based, data-intensive decision making such as in expert system applications. We suggest to selectively materialize the rule-generated data in relations so that rule-based decisions can be made incrementally and automatically when the collected data are updated, An algorithm is developed to select derived relations for materialization so that the overall cost of processing the inference rules is minimized while satisfying requirements on query response time. 1., |
