Post-Doctoral Research Visit in Neurosymbolic Relational Graph Neural Networks

The project aims to advance relational graph neural networks by incorporating symbolic knowledge into their message passing structures. This approach seeks to enhance the models' accuracy, robustness, and interpretability while reducing the need for large labeled datasets. The successful candidate will contribute to developing neurosymbolic extensions and benchmarks for relational learning, with results expected to be published in top-tier AI and database venues.