Theses

Information about writing student theses at DICE

Master

Robustness Evaluation of KG-augmented LLMs under Adversarial Attacks An Embedding Space for path-based Fact Checking Towards Meaningful Characteristic Graph Expressions A Dynamic Datastructure for Worst-Case Optimal Joins in Graph Databases Gradient-guided Refinement of Synthetic Knowledge Graphs

Bachelor

Byte pair encoding for Knowledge Graph Embeddings Benchmarking Linked Data processing systems A Dynamic Datastructure for Worst-Case Optimal Joins in Graph Databases Dynamic Label Relaxation RAG over Neural Triple Stores Stochastic Weight Averaging over Dynamic Graphs Simplifying OWL Class Expressions OWL Class Expression Learning over Tabular Data LLM in C over Knowledge Graphs SameAs Retrieval Service Prompt Engineered LLMs for Neural Link Prediction