BACKGROUND: Research on Artificial Intelligence in Education (AIEd) has become increasingly active during the past 20 years, with great interest, rich literature, and huge diversity. The diversity in research topics and technologies keeps increasing along with the tremendous growth in the application scope of AIEd research. A comprehensive overview of this field is necessary to understand what had been concerned in AIEd research, what is going on in the field of AIEd, and what might be the future of AIEd research. METHODS: This paper combined the structural topic modeling (STM) with the bibliometric analysis to automatically identify prominent research topics from the large-scale AIEd literature in the past two decades. Specifically, by using STM, we uncovered important research topics concerned within the AIEd community. The evolution of these topics was explored by using a nonparametric Mann-Kendall (MK) trend test. In addition, the correlation between these topics was further explored and visualized based on a semi-parametric Gaussian procedure. RESULTS: The annual trend of publications concerning AIEd had grown consistently across the past 20 years, indicating that research on AIEd had received a growing interest from academia Analyses on topical trends, correlations, and clusters reveal distinct developmental trends of these topics and promising research orientations. Specifically, the top six topics having the highest proportions in the dataset were cognition and perception, programming education, prediction, MOOCs, semantic Web and recommender systems, assessment and feedback, and emotion detection. Results of the MK test highlighted five topics receiving an increase in research interest, including prediction, assessment and feedback, flipped classroom and biosignal data, STEM education, and English language learning. In addition, several inter-topic research directions were identified, including 1) flipped classroom and biosignal data, emotion detection, and