Enrico Santi

Ph.D XL

Supervisor: Andrea Formisano, co-sup: Agostino Dovier

Phone:

Room: RIZ1 - L2c-09-BC

Research Project

My research focuses on Symbolic AI, specifically on the development of more scalable tools for inductive logic programming (ILP). ILP is an AI technique which can be used to generate explainable models. The explainability is intrinsic to the nature of the learned logic programs, which represent an interpretable knowledge base and inference rules.

These models are adaptable to various real-world tasks, functioning independently or alongside other systems. One notable application is in weather forecasting, where ILP models can enhance the interpretability of predictions generated by deep learning models, bridging the gap between symbolic and subsymbolic approaches.

A crucial aspect of the project involves leveraging GPGPU technology and developing CUDA-based parallel tools to accelerate ILP processes. This would enable faster learning of logic programs, making these explainable approaches more practical and accessible for broader use.

Support