The development of LLM tools for generating educational animations to enhance learning in mathematics
Educational videos have become a crucial content-delivery tool in all levels of education, particularly in blended and online learning environments. Creating high quality educational animations for mathematics instruction remains resource-intensive and technically challenging, limiting educator...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project / Dissertation / Thesis |
| Published: |
2025
|
| Subjects: | |
| Online Access: | http://eprints.utar.edu.my/7098/1/fyp_CS_2025_CEHL.pdf http://eprints.utar.edu.my/7098/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Educational videos have become a crucial content-delivery tool in all levels of
education, particularly in blended and online learning environments. Creating high
quality educational animations for mathematics instruction remains resource-intensive
and technically challenging, limiting educators' ability to produce engaging visual
content for blended and online learning environments. Traditional animation workflows
require specialized technical skills and substantial time investment, creating barriers
between pedagogical expertise and effective content delivery. This project introduces a
novel Domain-Specific Language (DSL) that enables Large Language Models (LLMs)
to automatically convert natural language descriptions of mathematical concepts into
structured animation commands. The core innovation lies in bridging natural language
input with precise animation instructions, allowing educators to generate visualizations
for complex topics such as function graphing, geometric proofs, and calculus
derivatives using simple text descriptions. The system architecture integrates GPT-4o
for intelligent content interpretation, a custom TypeScript animation library supporting
mathematical visualizations and transformations, and Google Text-to-Speech for
synchronized narration. Users input high-level prompts describing mathematical
concepts, which the LLM processes through the DSL to generate modular, reusable
animation scripts that render as web-based visualizations. This work contributes a
reproducible framework for AI-driven educational content creation that democratizes
animation production for mathematics instruction. The modular architecture and open
DSL design provide immediate practical value for mathematics educators while
establishing scalable foundations for broader STEM disciplines. The system represents
a significant advancement in the intersection of artificial intelligence, computer
graphics, and educational technology. By lowering the barriers to educational
animation creation, this project aims to enhance the learning experience for
mathematics students, with potential extension to broader STEM discipdlines. |
|---|