Enlarged Education – Exploring the Use of Generative AI to Support Lecturing in Higher Education. Hennekeuser, D., Vaziri, D. D., Golchinfar, D., Schreiber, D., & Stevens, G. International Journal of Artificial Intelligence in Education, August, 2024.
Enlarged Education – Exploring the Use of Generative AI to Support Lecturing in Higher Education [link]Paper  doi  abstract   bibtex   
Large Language Models (LLMs) are rapidly gaining attention across the open-source and commercial fields, bolstered by their constantly growing capabilities. While such models have a vast array of applications, their integration into higher education—as supportive tools for lecturers—has been largely unexplored. Exploring this area entails understanding the specific requirements and viewpoints of higher education lecturers. We developed an LLM-based assistant with retrieval augmented generation (RAG) capabilities and lecturing materials as its data foundation. For the design of the system, we followed a user-centered design approach. Subsequently, we conducted user studies and qualitative interviews with university lecturers. Our findings suggest that lecturers are ready to use LLMs with RAG in higher education under the condition that such systems are reliable, explainable, controllable, and trustworthy. We discuss design implications that designers of LLM-based systems should consider when developing such tools for higher education. This paper adds to the scarce existing studies on the usage of LLMs in educational contexts.
@article{hennekeuser_enlarged_2024,
	title = {Enlarged {Education} – {Exploring} the {Use} of {Generative} {AI} to {Support} {Lecturing} in {Higher} {Education}},
	issn = {1560-4306},
	url = {https://doi.org/10.1007/s40593-024-00424-y},
	doi = {10.1007/s40593-024-00424-y},
	abstract = {Large Language Models (LLMs) are rapidly gaining attention across the open-source and commercial fields, bolstered by their constantly growing capabilities. While such models have a vast array of applications, their integration into higher education—as supportive tools for lecturers—has been largely unexplored. Exploring this area entails understanding the specific requirements and viewpoints of higher education lecturers. We developed an LLM-based assistant with retrieval augmented generation (RAG) capabilities and lecturing materials as its data foundation. For the design of the system, we followed a user-centered design approach. Subsequently, we conducted user studies and qualitative interviews with university lecturers. Our findings suggest that lecturers are ready to use LLMs with RAG in higher education under the condition that such systems are reliable, explainable, controllable, and trustworthy. We discuss design implications that designers of LLM-based systems should consider when developing such tools for higher education. This paper adds to the scarce existing studies on the usage of LLMs in educational contexts.},
	language = {en},
	urldate = {2024-08-08},
	journal = {International Journal of Artificial Intelligence in Education},
	author = {Hennekeuser, Darius and Vaziri, Daryoush Daniel and Golchinfar, David and Schreiber, Dirk and Stevens, Gunnar},
	month = aug,
	year = {2024},
	keywords = {AI, Education, Human-centered Design, Large Language Models, Natural Language Processing, Semantic Search},
}

Downloads: 0