Mathematical modeling and mining real-world Big education datasets with application to curriculum mapping

Area of Science:

• Educational Data Mining
• Artificial Intelligence in Education

Background:

• University curricula present complex data structures.
• Identifying equivalent subjects across institutions is challenging for students.
• Developing personalized study plans requires effective curriculum data mining.

Purpose of the Study:

• To propose a novel approach for modeling and mining curriculum Big Data.
• To address the challenge of subject equivalency across different universities.
• To facilitate the creation of optimized, time-saving student learning paths.

Main Methods:

• Utilized concept graph-based learning techniques for curriculum data.
• Employed Bag of Words representations and natural language processing for subject descriptions.
• Developed a model to summarize subject networks and extract concepts automatically.

Main Results:

• Created ground truth for subject relations and descriptions.
• Successfully trained a model for subject network and concept graph generation.
• Validated the approach on nineteen real-world Australian university datasets.

Conclusions:

• The proposed approach effectively models and mines curriculum Big Data.
• Concept graph-based learning aids in distinguishing similar subjects across universities.
• The method provides a foundation for personalized and efficient student study planning.