Using reverse engineering techniques to infer a system use case model

Abstract

Summary In this article, a strategy that seeks to assist the arduous cognitive process that involves understanding a GUI‐based system is presented. In order to reach this goal, a UML use case model with the most relevant features is obtained. To derive this model, the strategy performs the following steps: Extraction and filtering of specific system static information and implementation of a clustering process that inspects this information including GUI widgets, which are components closely related to the system problem domain. Although these steps are commonly known and used in the context of reverse engineering, the strategy introduces unusual approaches regarding the proposals found in the available literature. More specifically, the strategy presents (a) a set of metrics that infers the relative importance of a method or a function within the analysed system, (b) a summarization process driven by different features of software systems, and (c) a technique to cluster software artefacts and to map the cluster model onto a use case model. The article also propose a methodology that allows to compare the achieved results. The assessment of the approach suggests that the strategy can assist the software engineer to understand a software system providing a useful fine‐grained use case model.

Reversely Discovering and Modifying Properties Based on Active Deep Q-Learning

“…Poženel [3] used clickstream data to enhance reverse engineering of Web applications. Miranda [4] used reverse engineering techniques to infer a system use case model. It is possible to reversely generate a three-dimensional model from a set of 2D photographs [5].…”