Towards Interactive Learning for Model-based Software Engineering

Abstract

Model-based technologies improve the efficiency of the design and development of IT systems by making it possible to automate verification, code generation and system analysis based on a formal model. A simple way of describing the behavior of systems is state-based modeling, which - due to the advancements of formal analysis techniques in recent years - can be widely and effectively utilized when analyzing systems. A possible way of synthesizing such models is to apply active automata learning algorithms. Acquiring a correct formal model of a system can be challenging because of the different theoretical and practical obstacles of both manual and automated approaches. We propose a semi-automated solution, that applies automata learning to provide an interactive environment for model development.