Automated, traceable, and interactive domain modelling

Abstract

In domain modelling, practitioners manually extract analyzable and more concise domain models from problem descriptions which express requirements in natural language. With automated domain modelling support using existing approaches, some challenges remain unaddressed - inadequate accuracy of extracted models, no support for traceability of modelling decisions, and no facility for system-modeller interactions. To address these challenges and better support practitioners, we present our bot-assisted solution. Furthermore, we evaluate the effectiveness of our solution and find promising results which warrant further research in this direction.