Descriptive Modeling of Collaborative Practices in a Computer-Based Human Learning Environment
No Thumbnail Available
Date
2025
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In an increasingly digital educational context, Computer-Based Human Learning Environments
(CBHLEs) play a crucial role in supporting both individual and collaborative learning.
Collaborative learning supported by these environments offers a major opportunity to
strengthen learner engagement and improve the quality of peer interactions. However, to
fully leverage this potential, it is essential to understand and assess learning dynamics using
reliable indicators derived from interaction trace analysis.
This thesis focuses on the design and computation of collaboration indicators in ComputerBased
Human Learning Environments (CBHLEs). It proposes a model-driven architecture
(MDA)-based approach to make the indicator computation platform-independent.
Three main contributions are presented:
• First contribution: a model transformation-based approach is proposed to compute
collaboration indicators independently of the learning platform.
• Second contribution: a formal model (DECIN Model) is introduced to assist teachers
in designing valid collaboration indicators tailored to their observation needs.
• Third contribution: the computation process is automated through the generation of
Sequences of transformations, generated by a formal system (DECIN-AGST System).
To support this approach, a platform-independent trace model (T-PIM) is proposed,
along with formal constraints expressed in OCL to ensure the validity and consistency of the
collected traces. Similarly, a generic model for collaborative indicators (CI-PIM) is defined,
supported by constraints that ensure the quality of the generated indicators.
Finally, the automation of transformation sequence generation is achieved through the
implementation of the DECIN model in a formal system (DECIN-AGST), whose validity has
been verified using the model checking method via the UPPAAL tool, thus ensuring safety,
accessibility, and absence of deadlocks.