A framework for designing and analyzing multimodal learning analytics systems

• The integration of technology in education offers transformative potential, especially with the advent of data-driven approaches that can personalize learning, support educators, and provide valuable insights into the learning process. Multimodal learning analytics (MMLA) holds remarkable promise within this context. By capturing and analyzing data from multiple sources—including video, audio, and digital interactions—MMLA systems offer a holistic view of learning experiences and the ability to tailor interventions in real time. This application has profound implications for understanding and enhancing learning experiences. However, the design of such sophisticated systems poses a significant challenge. Without conventional and field-tested frameworks, MMLA system development often remains self-driven and tailored to specific contexts, limiting both these systems’ broader adoption and full utilization. This thesis proposes a structured framework for designing MMLA systems across diverse educational contexts to address this fundamental challenge. The development of the framework followed a multifaceted methodology. In addition, action design research involving empirical studies, literature reviews, and expert interviews was employed to establish a set of foundational design considerations. The framework was then applied and refined within real-world educational settings. These included applications in the context of a globally distributed classroom and language acquisition environments. This practical application led to refinements that enhanced the framework’s adaptability and user-centric design. This thesis makes three key contributions: (1) a set of design considerations for MMLA systems, (2) a framework offering a structured guide for the design of MMLA systems, and (3) a conceptual system demonstrating the framework’s principles. The implications of this work are significant for researchers and stakeholders in MMLA, providing a foundation for future MMLA system development and ensuring more systematic and conventional design practices. This structured approach paves the way for broader adoption and integration of MMLA, ultimately enhancing educational outcomes and fostering personalized learning environments.

Ämnesord

NATURVETENSKAP  -- Data- och informationsvetenskap (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences (hsv//eng)

Nyckelord

Multimodal Learning Analytics

Educational Technology

Smart learning Environments

Internet of Things

Publikations- och innehållstyp

vet (ämneskategori)

dok (ämneskategori)