The reflective approach for data-driven rules

Abstract

The self-triggering approach presented in this paper is a novel architecture for active database systems. It provides a high-level reflective rule language, rich modelling tools and a clear semantics of dependencies among data elements. Under this approach, rules that are data-driven by nature are handled by a mechanism that utilizes their semantic properties, while rules that are event-driven by nature are handled with extended modelling facilities. The self-triggering approach has been employed by the PARDES project that is described throughout this paper. This approach is contrasted with the "event-control-action" method, the leading architecture in current active database models.<>