"Discovery Event-driven Action and Data Modelling"

Supervised by A.Simons

Abstract

The aim is to construct a CASE tool for the forthcoming version 2 of the Discovery Method that follows an event-driven approach to derive the passive entities necessary for data storage and the manager-entities for handling the coordination of messages. The approach is based on 'Event-driven-design' (Simons and Snoeck), where the events are pulled from a state domain until all objects are linked in a compositional hierarchy and the chain of responsibility is established. The CASE tool should accept a Task Model and narratives as input; from this is should derive an object-event table and build the composition graph, by identifying subsets of objects that are involved in the same set of events.

This thesis takes from a thesis currently being developed by David Carrington (titled CASToff). His thesis works on the initial development of task structure and flow, providing the groundwork for this thesis, which takes the output of Carrington's project and manipulates this to produce the state model and associated object-event table and existence-dependency graph. Though not complete, references will be made to Carrington's work and its association with this project throughout.