Mapping Artificial Intelligence in Education Research: a Network‐based Keyword Analysis

Publication Information


  • Shihui Feng, The University of Hong Kong
  • Nancy Law, The University of Hong Kong


  • 277-303


  • Artificial intelligence in education, Network analysis, Keyword co‐occurrence network, Knowledge mapping, Systematic review


  • In this study, we review 1830 research articles on artificial intelligence in education (AIED), with the aim of providing a holistic picture of the knowledge evolution in this interdisciplinary research field from 2010 to 2019. A novel three-step approach in the analysis of the keyword co-occurrence networks (KCN) is proposed to identify the knowledge structure, knowledge clusters and trending keywords within AIED over time. The results reveal considerable research diversity in the AIED field, centering around two sustained themes: intelligent tutoring systems (2010-19) and massive open online courses (since 2014). The focal educational concerns reflected in AIED research are: (1) online learning; (2) game-based learning; (3) collaborative learning; (4) assessment; (5) affect; (6) engagement; and (7) learning design. The highly connected keywords relevant to analytic techniques within this field include natural language processing, educational data mining, learning analytics and machine learning. Neural network, deep learning, eye tracking, and personalized learning are trending keywords in this field as they have emerged with key structural roles in the latest two-year period analyzed. This is the first article providing a systematic review of a large body of literature on artificial intelligence in education, and in it we uncover the underlying patterns of knowledge connectivity within the field, as well as provide insight into its future development. The three-step multi-scale (macro, meso, micro) framework proposed in this study can also be applied to map the knowledge development in other scientific research areas.