CourseGraph: Automated Evaluation of Course Overlaps for Student Mobility
A novel approach named CourseGraph has been introduced to facilitate the assessment of external courses for student mobility initiatives such as Erasmus+. This system retrieves course details from online sources, semantically encodes the information using a BERT-based language model, and employs a Random Forest classifier to assess the compatibility of foreign courses with those in the home curriculum. The goal is to simplify the workload for curriculum administrators, preventing students from enrolling in courses that significantly overlap with their existing programs. This methodology is elaborated upon in a paper available on arXiv (2608.05910).
Key facts
- CourseGraph is a methodology for automated evaluation of external courses.
- It is designed for student mobility programs such as Erasmus+.
- The system extracts course titles, descriptions, and learning outcomes from course webpages.
- It uses a BERT-based language model for semantic representation.
- Pair-wise similarity between courses is computed.
- A Random Forest classifier determines overlap.
- The paper is available on arXiv with ID 2608.05910.
- The methodology is based on insights from curriculum administrators.
Entities
Institutions
- arXiv