Doctoral studies

Doctoral thesis studied how to support the development of context-aware and reusable Multimodal Learning Analytics solutions

Shashi Kant Shankar

To support evidence-based teaching and learning practices in authentic learning settings, Multimodal Learning Analytics (MMLA) can play an important role. Therefore, context-aware and reusable MMLA solutions should be developed and adopted in our daily formal educational practices, believes Shashi Kant Shankar, who defended his doctoral thesis at Tallinn University.

Recent advancements and the wide reach of digital technologies, network communication, artificial intelligence, and data science have enabled us to adopt evidence-based practices in our daily lives. Educational practices like teaching, learning, curriculum development, and assessment are not an exception to this list. For example, traditional Learning Analytics (LA) is harnessing the digital logs of Learning Management Systems since 2011. Similarly, Multimodal Learning Analytics (MMLA) is trying to harness multimodal evidence of learning collected from physical as well as digital space of a learning situation. Because of this, Shashi’s doctoral thesis focused on supporting the development of MMLA solutions that can be adopted in authentic learning settings.

Based on his research, there is a need for context-aware and reusable MMLA solutions so that MMLA as an approach can be adopted. However, the development of context-aware and reusable solutions is complicated because multiple cross-disciplinary stakeholders need to be involved in requirements specification and documentation, the contextual information of a learning situation is not structured and organized, technical complexity in processing multiple heterogeneous data sources, and ad-hoc and opportunistic software development practice. Shankar’s thesis contributes an MMLA infrastructure that involves a data value chain to support the cross-disciplinary stakeholders’ communication while specifying requirements, a data model to structure and organize design- and enactment-related contextual information, and a software architecture.

The doctoral thesis contributed to the MMLA knowledge base in terms of requirements specification, context awareness, and reusability of MMLA solutions. The doctoral thesis also presents guidelines for teachers, researchers, and developers for planning an MMLA scenario, promoting the adoption of MMLA, and supporting the reusability of MMLA solutions in authentic learning settings.

Tallinn University School of Digital Technoligies new doctor Shashi Kant Shankar defended his doctoral thesis “CIMLA: a data infrastructure to support the development of context-aware and reusable multimodal learning analytics solutions” on 21st March 2023. Supervisors were Tallinn University associate professors Luis P. Prieto and María Jesús Rodríguez-Triana and advisor Adolfo Ruiz-Calleja, a senior researcher at the University of Valladolid. Opponents were University of Valladolid professor Juan Ignacio Asensio Pérez and University College London professor Mutlu Cukurova.