ABSTRACT

In order to maintain a system, it’s critical to understand its architecture. However even though every system has an architecture, not every system has a reliable representation of its architecture. To deal with this problem many researchers have engaged in software architecture extraction where the system’s architecture is recovered from its source code. While there is a plethora of approaches aiming at extracting software architectures, there is no mean or tool measurement for these approaches; which makes the comparison between the different approaches a hard task. To tackle this lack, we developed a meta-model, based on SPEM meta-model, that specifies the software architecture extraction process. Such meta-model serves as a tool to compare, analyze and evaluate research field approaches. In this paper we detail our meta-model called SArEM (Software Architecture Extraction Meta-model) and clarify its concepts.