Procedural Content Generation for Math Education

Master Thesis (2020)
Author(s)

Y. Xu (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Contributor(s)

Rafael Bidarra – Mentor (TU Delft - Computer Graphics and Visualisation)

E. Eisemann – Graduation committee member (TU Delft - Computer Graphics and Visualisation)

Christoph Lofi – Graduation committee member (TU Delft - Web Information Systems)

Faculty
Electrical Engineering, Mathematics and Computer Science
Copyright
© 2020 Y. Xu
More Info
expand_more
Publication Year
2020
Language
English
Copyright
© 2020 Y. Xu
Graduation Date
14-08-2020
Awarding Institution
Delft University of Technology
Programme
['Computer Science']
Faculty
Electrical Engineering, Mathematics and Computer Science
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Mathematics education plays an essential role in children’s development. In the past few years, online mathematics learning has gained increasing popularity. The online learning platform needs a large variety of textual and visual content to offer children a convenient learning experience and help them practice various mathematical skills. However, manually creating content is hugely time-consuming, expensive, and tedious for the content editors.

This project proposes a generic approach for procedural generation of mathematical problems and corresponding textual and visual content. We analyzed and clustered hundreds of primary school curriculum-based math knowledge components, and built flexible templates for generating abstract math problems, including arithmetic, comparison, ordering, mathematical relationships, measurements, and geometry. Then our system realizes the abstract math problems in natural language through the lexicalization of language-independent semantic configurations and syntactically structured templates. Our system generates visual content through text-based image retrieval and visualization of abstract math content, varying in the forms of table, chart, geometry, or picture for counting objects. Human expert evaluations found that our generated contents are understandable, sensible, and achieve well usefulness for primary school students.

Files

Final_Thesis_Report_Yi.pdf
(pdf | 6.72 Mb)
License info not available