The focus of the proposed thesis is to design a framework based on graph neural networks that can deal with multi-relational knowledge graphs, such as DBpedia and Wikidata. The student will implement, evaluate, and optimize the framework on current datasets and compare the performance with the state-of-the-art in new application areas (e.g., information extraction from text). Initial code and data will be provided to the student.
Further information can be found in the attached PDF.
Ausschreibung: Download (pdf)