Knowledge Tracing and Learner Characteristic Analysis in Intelligent Education: A Bibliometric Study of Trends, Themes, and Emerging Directions

Abstract

Knowledge Tracing (KT) and learner characteristics have become key research areas in intelligent education, aiming to model and predict students’ learning processes over time. Despite the rapid growth of this field, a comprehensive understanding of its scientific structure and long-term evolution remains limited. To address this gap, this study presents a large-scale bibliometric analysis of KT research based on 688 publications indexed in the Scopus database between 2010 and 2026. The results reveal a sustained and accelerating growth in scientific production, particularly after 2020, driven by the increasing integration of advanced artificial intelligence techniques. Thematic and temporal analyses highlight a clear transition from traditional probabilistic models, such as Bayesian Knowledge Tracing, toward data-driven approaches, including deep learning, graph-based models, and privacy-preserving methods such as federated learning. In addition, the keyword co-occurrence network reveals a highly interconnected research structure, where core concepts such as knowledge tracing, student modeling, and learning systems act as central nodes linking educational applications with machine learning methodologies. The analysis was conducted using the bibliometrix $\mathbf{R}$ package and VOSviewer, enabling a comprehensive exploration of publication trends, collaboration patterns, and thematic evolution. This study contributes by providing a structured and quantitative overview of the KT research landscape, identifying its main research streams, and highlighting emerging directions for future work in intelligent and adaptive educational systems.