Regresar

Model-driven development of multi-agent systems: A systematic mapping

Abstract:

Context: Model-Driven Development (MDD) and Multi-Agent Systems (MAS) are two challenging research areas in software engineering. MDD is about improving software development whilst Multi-Agent Systems (MASs) are systems in which many intelligent autonomous entities (agents) interact with each other for solving their selfish or common problems, and to achieve certain goal. Hence examining the available proposals in the literature to use MDD when developing MAS may be insightful. Objective: This work aims at assessing the state of the art in MDD for MAS. It mainly focuses on: what are the characteristics of MDD approaches that support MAS; what types of MAS are supported and what is the impact of these approaches in the quality of both the development process and system. Method: We carried out a mapping study following a formal guide. We applied a search string over the set of three popular browsers (Scopus, IEEE and ACM). Consequently, 78 papers were selected and analyzed with respect to the defined classification criteria derived from the research questions. Results: The analysis allows us to answer the research questions. We remark: 1) predominance of papers from academic environment and written by researchers only; 2) predominance of top-down transformation in software development activities; 3) inexistence of consolidated methods; 4) significant percentage of works without tool support; 5) wide variety of modeling notation and languages; 6) lack of empirical evidences to demonstrate the actual impact of the approaches; 8) limited application in real cases. Conclusion: This study provides the state of the art in the topic and identifies problems that deserve investigations in the future, for instance the significant percentage of works without tool support or the lack of real case studies.