Profile

Samy received his bachelor’s in civil engineering and a master’s in computer science and engineering both from the Free University of Brussels. During his master studies, he joined Sony Computer Science Laboratories for a three-month internship and fell in love with research. He decided to join Sony AI after graduating to grow further in this environment. His research focus is in neurosymbolic AI, where he explores non-classical logical operators and their mathematical properties in machine learning contexts.

Message

“My current focus at Sony AI is in the Gastronomy project and my expertise is in neurosymbolic AI. Neurosymbolic AI is a field that combines reasoning from complex knowledge representations, such as logical knowledge bases or semantic networks, with learning from data. I thrive to address topical industry problems such as understanding gastronomy and health by relying on machine learning capabilities and pre-existing domain and logical knowledge. When there are ways to efficiently let a neural network know that cats are mammals, without learning it from scratch, I want to explore them.”

Publications

What's Wrong with Gradient-based Complex Query Answering?

NeSy, 2023
Ouns El Harzli, Samy Badreddine, Tarek Besold

Multi-hop query answering on knowledge graphs is known to be a challenging computational task. Neurosymbolic approaches using neural link predictors have shown promising results but are still outperformed by combinatorial optimization methods on several benchmarks, including…

Logic Tensor Networks

Artificial Intelligence, 2022
Samy Badreddine, Artur d'Avila Garcez*, Luciano Serafini*, Michael Spranger

Attempts at combining logic and neural networks into neurosymbolic approaches have been on the increase in recent years. In a neurosymbolic system, symbolic knowledge assists deep learning, which typically uses a sub-symbolic distributed representation, to learn and reason a…

Extending Real Logic with Aggregate Functions

IJCLR, 2021
Samy Badreddine, Michael Spranger

Real Logic is a recently introduced first-order language where formulas have fuzzy truth values in the interval [0, 1] and semantics are defined concretely with real domains. The Logic Tensor Networks (LTN) framework has applied Real Logic to many important AI tasks through …