Inferring Use-cases from GUI Analysis

Abstract

One of the most arduous and tedious tasks in the life cycle of an application is Software Maintenance and Evolution. In this context, the most time-consuming activities are those that the programmer must execute to get a complete understanding of the system. Based on this need, Program Comprehension (PC), a Software Engineering discipline, arises to tackle the problem. This article presents SSPIA, a strategy that assists software engineers to understand a system by inferring its use case model. SSPIA uses concepts, techniques and tools of PC to carry out its purposes. In order to extract use cases, some specific system static information is collected. This information serves as input to a process that implements a clustering technique based on systems Graphical User Interface (GUI). Almost all clustering techniques take as main criteria graphs structural properties. However, in the context of PC, some Problem Domain information must be considered. The strategy proposed in this article takes into account an essential component strongly related to systems Problem Domain: Graphic User Interfaces. As a main aim, the proposed strategy seeks to assist the arduous cognitive process that involves understanding a system.