Investigating Open-Source Large Language Models in Digital Pedagogies

Abstract

As higher education continues to embrace digital pedagogies, large language models (LLMs) present opportunities for improving student-centered learning. Open-source LLMs make advanced AI technology available to a wide range of researchers, developers, and organizations and can be adapted and fine-tuned for specific tasks or domains, allowing for more specialized and resource-efficient educational applications. Higher education often involves more self-directed and autonomous learning, especially in online and distributed learning environments. Additionally, since higher education incorporates more advanced and specialized topics, AI systems must support complex subject matter and facilitate interdisciplinary connections. This study examines the integration of open-source LLMs into digital pedagogical frameworks to promote critical thinking, collaboration, and self-regulation in higher education contexts. The study employs a scoping review approach to map existing literature and current implementations, identifying key concepts and gaps in the research community. Furthermore, it investigates effective training methods for educators to ensure these tools are implemented to their fullest pedagogical potential. Through the analysis of case studies and current practices, the study demonstrates the transformative impact of LLMs in creating more inclusive and responsive educational experiences. The findings from this study provide key insights into how higher education institutions can better integrate LLMs, offering a framework for future research on the development of AI-driven educational tools.

Presenters

Saeed Saffari
Student, Master, Dalhousie University, Nova Scotia, Canada

Jeeho Ryoo
Assistant Professor, Fairleigh Dickinson University, Canada

Oscar Lin
Prof., School of Computing and Information Systems, Athabasca University, Alberta, Canada

Michael Lin
Assistant Professor, Faculty of Education, Mount Saint Vincent University, Nova Scotia, Canada

Details

Presentation Type

Paper Presentation in a Themed Session

Theme

Considering Digital Pedagogies

KEYWORDS

Open-Source Large Language Models, Digital Pedagogies, Artificial Intelligence in Education