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

    Abstract

    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
    Pages338-342
    Number of pages5
    DOIs
    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

    Conference

    Conference23rd International Conference on Artificial Intelligence in Education, AIED 2022
    Country/TerritoryUnited Kingdom
    CityDurham
    Period27/07/2231/07/22

    Keywords

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

    Fingerprint

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

    Cite this