Research & Papers

RAGEAR: Retrieval-Augmented Graph Boosts Academic Course Recommendations

Combines dense transcript search with a knowledge graph for smarter course picks.

Deep Dive

RAGEAR is a neurosymbolic recommender system for academic courses that combines dense retrieval over full lecture transcripts with a symbolic knowledge graph modeling courses, lessons, transcript chunks, credits, study plans, and curricular information. Its key contribution is a graph-aware aggregation function that scores courses using three factors: the share of retrieved similarity for a course, the rank-based strength of its relevant chunks, and the distribution of evidence across lessons. Evaluated on 152 student-like queries, results show that lecture transcripts improve over metadata-only retrieval and that RAGEAR further improves ranking quality over a transcript-based normalized SumP baseline, especially for top-ranked recommendations.

Key Points
  • RAGEAR performs dense retrieval over full lecture transcripts, not just metadata, to capture fine-grained course content.
  • It uses a symbolic knowledge graph to filter and contextualize results based on credits, prerequisites, and study plans.
  • Evaluated on 152 student queries with both human judges and an LLM—outperformed baselines, especially for top-ranked courses.

Why It Matters

Smarter, more transparent course recommendations that combine semantic search with academic constraints—reducing student guesswork.