Towards a Framework for Self-adaptive Systems Conceptual Modeling

Abstract

Context: Self-adaptive Systems (SaSs) requirements engineering is challenging because it is necessary to deal with invariant requirements, which specify the system's crisp behavior, and adaptive requirements, which specify the system's fuzzy behavior. Problem: Our motivational question is: how to make the process of adaptive requirements conceptual modeling less complex, ensuring the coverage of the main abstractions and the comprehensibility of the resultant model? Solution: To address this, we propose a modeling framework that defines a strategy for adaptive requirements conceptual modeling based on a set of design patterns. IS Theory: This work is proposed based on the adaptive structuration theory. Method: The proposed framework was made by analyzing the main concepts related to SaSs, finding a suitable set of design patterns, specifying a modeling schema with UML, and evaluating it in focus group sessions. Summary of Results: The focus group results point out that the proposed framework supports the adaptive requirements conceptual modeling, guiding the modeling process, mainly to non-expert developers. Contributions and Impact in the IS area: Moreover, we highlight some relevant contributions of the framework, it is able to provide support for non-expert developers by guiding the SaSs conceptual modeling process. It also has an abstract modeling schema that can be applied to other concrete scenarios. Furthermore, its well-formed structures make it more comprehensible and didactic to stakeholders.