An ontology-based approach for integrating tools supporting the software measurement process

Abstract

Context: Software measurement is a fundamental practice to support process improvement and project management, since it provides useful data for decision making at both organizational and project levels. Due to the nature of software measurement activities, the use of computational supporting tools is essential. Software measurement can be performed in the context of various software processes and these processes generally have different supporting tools. Thus, it is common to use several tools to allow collecting data regarding the processes. Tools are usually developed at different time, by different teams and without concern for integration. As a result, organizations have to deal with integration issues to enable communication between tools and to properly support the measurement process. A key factor for integration is that tools share a common understanding regarding the meaning of the exchanged terms and services. In other words, it is important to deal with integration not only at the syntactic level, but also at the semantic level. Among the instruments used to address semantics, ontologies have been acknowledged as an important means to address semantic integration. Objective: This paper presents the Ontology-Based Approach for Measurement Systems Integration (OBA-MSI), an approach that uses ontologies as a basis to integrate tools aiming at supporting the software measurement process. Method: OBA-MSI was developed following the Design Science Research paradigm. To evaluate OBA-MSI, we carried out a case study in which we used the approach to integrate tools for a real software development organization. After that, we applied a survey to get feedback from users of the integrated solution. Results: The users stated that the integrated solution obtained from applying OBA-MSI properly supported the software measurement process and added value to the organization, providing more benefits than using the tools in isolation. Conclusions: The use of OBA-MSI to aid tools integration helps deal with semantic conflicts and contributes to obtain a proper support to the software measurement process.