Learning Optimal and Personalized Knowledge Component Sequencing Policies

Fuhua Lin, Leo Howard, Hongxin Yan

Research output: Chapter in Book/Report/Conference proceedingPublished Conference contributionpeer-review

1 Citation (Scopus)


One of the goals of adaptive learning systems is to realize adaptive learning sequencing by optimizing the order of learning materials to be presented to different learners. This paper proposes a novel approach to recommending optimal and personalized learning sequences for learners taking an online course based on the contextual bandit framework where the background knowledge of the learners is the context. To improve learning efficiency and performance of learners, the adaption engine of such an adaptive learning system can select an optimal learning path for a learner by continually evaluating the learners’ progress as the course advances. To overcome the complexity of learning path recommendation due to the large number of knowledge components, we use the ‘divide-and-conquer’ approach to modeling the domain and designing the sequence adaptation algorithm. Also, the adaptation engine can dynamically replan the learning path for a learner if her/his performance is worse than expected. Finally, our approach can improve over time by learning from the experience of previous learners who adopted recommended sequences.

Original languageEnglish
Title of host publicationArtificial Intelligence in Education. Posters and Late Breaking Results, Workshops and Tutorials, Industry and Innovation Tracks, Practitioners’ and Doctoral Consortium - 23rd International Conference, AIED 2022, Proceedings
EditorsMaria Mercedes Rodrigo, Noburu Matsuda, Alexandra I. Cristea, Vania Dimitrova
Number of pages5
Publication statusPublished - 2022
Event23rd International Conference on Artificial Intelligence in Education, AIED 2022 - Durham, United Kingdom
Duration: 27 Jul. 202231 Jul. 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13356 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference23rd International Conference on Artificial Intelligence in Education, AIED 2022
Country/TerritoryUnited Kingdom


  • Adaptive learning
  • Bandit algorithms
  • Learning path recommendation
  • Sequence adaptivity


Dive into the research topics of 'Learning Optimal and Personalized Knowledge Component Sequencing Policies'. Together they form a unique fingerprint.

Cite this