Interactive Situation Models for Program Comprehension
Meurig Beynon, Pi Sun · Warwick Research Archive Portal (University of Warwick) · 1998
The complexity of the interactions between programmable components and human agents in modern computing applications motivates new approaches to program comprehension. Understanding the role of programs within a reactive system, for instance, involves not only input-output transformations, but also communication and stimulus-response issues. This paper examines the prospects for constructing novel computer-based interactive situation models to assist program comprehension. Such a model provides an environment within which the human interpreter can explore the data relationships and patterns of behaviour generated by a computer program with particular reference to its external real-world semantics. The proposed method of construction exploits principles based upon observation, agency and dependency ("Empirical Modelling") that have been developed at the University of Warwick. Connections between these principles and previous work on program comprehension are discussed. Some experimental studies in applying interactive situation models are reviewed, and conclusions about their potential future role in program comprehension and generation are drawn.