Knowledge Graph Enhanced Recommender Systems
B.Sc. thesis (Aalto University, graded 5/5): a literature review of knowledge graph-enhanced recommender systems against collaborative, content-based and hybrid filtering.
B.Sc. Thesis · Aalto University, School of Science · September 2025 · Graded 5/5 (Excellent)
A literature review comparing traditional recommender systems (collaborative filtering, content-based filtering, hybrid methods) against knowledge graph-enhanced approaches (graph neural networks and knowledge graph embeddings). KG-based systems address the cold-start problem and improve explainability through path-based reasoning, at the cost of higher computational overhead and construction complexity.