An MDA-based Approach for the Design and Automatic Computation of Collaboration Indicators in E-Learning Systems

Effectively assessing collaboration in e-learning environments requires robust indicators that capture learner interactions. However, most existing approaches are platform-dependent, rely on predefined indicators, and offer little flexibility for tutors and teachers. This limits their ability to tailor assessments to specific learning contexts. To address these challenges, this study proposes a Model-Driven Architecture (MDA)-based approach that allows tutors and teachers to design and compute collaboration indicators independently of any specific platform, without requiring advanced technical expertise. This approach introduces a flexible and scalable framework for the design and computation of collaboration indicators in e-learning analytics. First, it ensures platform-independent computation by leveraging a model transformation approach within an MDA-based process. Second, it provides a formal model that enables teachers and tutors to define meaningful and adaptable indicators that align with their specific observation needs. Finally, the proposed DECIN-AGST system automates the generation of transformation sequences. The accuracy and reliability of the proposed system were validated through formal verification using the UPPAAL model checker.