Conceptualizations of Information Science by Large Language Models
Abstract
This paper reports a comparative study of how large language models understand and represent the domain of information science. Five large language models were selected for this study, namely ChatGPT, Perplexity.ai, Google Gemini, Meta AI and Claude. A set of five prompts was utilized in this study for comparison. The findings suggest differences and variations in how these LLMs conceptualize and represent information science, its definitions, and interdisciplinarity, theoretical models, and methods.
Ali Shiri
School of Library and Information studies, University of Alberta
Ali Shiri is a Professor in the Faculty of Education and is currently the Vice Dean of the Faculty of Graduate & Postdoctoral Studies. He received his PhD in Information Science from the University of Strathclyde Department of Computer and Information Sciences in Glasgow, Scotland. Ali has been teaching, researching, and writing about digital repositories, digital information interaction, data and learning analytics, and more recently on the generative AI implications and ethics in higher education and research. In his current research, funded by the Social Sciences and Humanities Research Council of Canada (SSHRC), he is developing mobile applications for cultural heritage digital libraries and digital storytelling systems for the Inuvialuit communities in the Northwest Territories in Canada’s Western Arctic.