The development of LLM tools for generating educational animations to enhance learning in mathematics

Chung, Elaine Hui Lin (2025) The development of LLM tools for generating educational animations to enhance learning in mathematics. Final Year Project, UTAR.

PDF Download (3192Kb)

Abstract

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.

Actions (login required)