Home // INTELLI 2017, The Sixth International Conference on Intelligent Systems and Applications // View article

Developing Space Efficient Techniques for Building POMDP Based Intelligent Tutoring Systems

Authors:
Fangju Wang

Keywords: Intelligent system; intelligent tutoring system; adaptive teaching; partially observable Markov decision process; space efficiency.

Abstract:
In building an intelligent tutoring system (ITS), the partially observable Markov decision process (POMDP) model provides useful tools to deal with uncertainties, which are major challenges in achieving adaptive teaching. However, the POMDP model is very expensive. When a method of policy trees is used in decision making, the number of trees and sizes of individual trees are typically exponential. The great space complexity obstructs application of the POMDP model to ITSs. In our research, we developed space efficient techniques to address the space complexity problem. The techniques minimize the number and sizes of trees, and reduce space consumption of the tree database. Encouraging results have been achieved: the techniques enabled us to build a system with a manageable size, to teach a practical subject.

Pages: 44 to 50

Copyright: Copyright (c) IARIA, 2017

Publication date: July 23, 2017

Published in: conference

ISSN: 2308-4065

ISBN: 978-1-61208-576-0

Location: Nice, France

Dates: from July 23, 2017 to July 27, 2017