Science Mapping of the Knowledge Base on the Effects of Artifcial Intelligence (AI)-Based Chatbots on Student Learning Outcomes: A Bibliometric Study

Abstract

With the advent of AI-based Technologies, several groundbreaking tools have become part of the digitalization process in various fields, and the education field was no exception. The growing emergence of several Artificial Intelligence (AI)-based chatbots that can engage with human-like thinking and performance particularly attracted the attention of educational researchers and practitioners with their strong capacity to support student learning. This new line of research has accumulated a large knowledge base on the influence of these chatbots on students’ learning outcomes in a very short time span. However, in order to lead the efficient growth of this knowledge base, a detailed analysis of its thematic architecture, intellectual evolution, and bibliometric performance is warranted. Therefore, the current study conducted a science mapping analysis of this field using the SciMAT software, which uniquely combines the analysis of the bibliometric performance and conceptual development, allowing for comparisons across several periods of analysis. Data for the current study was collected from the WoS and Scopus databases. The analysis of 322 articles showed that studies published during the first period of analysis (2011–2022) mostly focused on how AI chatbots could act as intelligent tutoring systems, particularly in online courses. During the second period (2023), though, this attention turned to ChatGPT and concerns over its influence on the integrity and academic honesty of students. During the last period of analysis (2024), investigations tended to move from concerns over its side effects to how these chatbots could support customized learning and students’ self-learning experiences. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.