My research interests primarily lie at the intersection between deep learning and knowledge representation and reasoning. In particular, I am interested in developing hybrid neuro-symbolic systems, based on neural networks and explicit logical structures, to more efficiently and reliably learn patterns and make inferences from data. The main aim of my research is to produce systems that combine the learning capacity of deep neural networks with the reliability, safety, and interpretability of logical systems.
I earned a B.E. in Computer Engineering from the Lebanese American University (LAU), with a minor in Mathematics, in 2017, and my capstone project, supervised by Dr. Joe Tekli, was on automated sentiment-aware music composition using machine learning and genetic algorithms. I then completed an MSc in Computer Science at the University of Oxford in 2018, with my thesis, supervised by Prof. Daniel Kroening, tackling neural program synthesis. Parallel to my studies, I also earned a Baccalaureate degree in Piano performance from the Lebanese National Higher Conservatory of Music (LNHCM) in 2017.
The Surprising Power of Graph Neural Networks with Random Node Initialization
Ralph Abboud‚ İsmail İlkan Ceylan‚ Martin Grohe and Thomas Lukasiewicz
In Proceedings of the 30th International Joint Conference on Artificial Intelligence‚ IJCAI 2021‚ August 21–26‚ 2021. IJCAI. August, 2021.
BoxE: A Box Embedding Model for Knowledge Base Completion
Ralph Abboud‚ İsmail İlkan Ceylan‚ Thomas Lukasiewicz and Tommaso Salvatori
In Proceedings of the 34th Annual Conference on Neural Information Processing Systems‚ NeurIPS 2020‚ December 6–12‚ 2020. December, 2020.
Neural Networks for Approximate DNF Counting: An Abridged Report
Ralph Abboud‚ İsmail İlkan Ceylan and Thomas Lukasiewicz
In Lingfei Wu‚ Jian Tang‚ Yinglong Xia and Charu Aggarwal, editors, Proceedings of the 1st International Workshop on Deep Learning on Graphs: Methodologies and Applications‚ DLGMA 2020‚ New York‚ New York‚ USA‚ February 8‚ 2020. February, 2020.