Representations and Reasoning with Uncertain Data and Knowledge

  • Leopoldo Bertossi – Universidad Adolfo Ibáñez (UAI), Santiago de Chile, Chile. Skema Business School, Montreal, Canadá.

Learning Goals

Relational representations of uncertain data and knowledge are crucial in artificial intelligence (AI) and machine learning (ML) when applied to richly structured domain models. Probabilistic Databases and Relational Probabilistic Graphical Models, that can be used to create those models and interact with them, share several problems or representation, reasoning and inference. Techniques for each of them can be applied to the other.

In this course this connection will be presented. In particular, “model counting» techniques will be unveiled and exploited. A central topic is that of capturing and representing domain knowledge and domain semantics. It is in this direction that Statistical Relational Learning has most to contribute to AI and ML prac-
tice and research, in particular in combination with deep learning techniques,in the emergent area of neuro-symbolic AI. This course will touch on all these topics, providing the basis for starting research in the area.


Probabilistic databases. Probabilistic graphical models. Statistical relational representations. Probabilistic logic programming. Probabilistic conceptual models and ontologies.


“Probabilistic Databases». Suciu, D., Olteanu, D., Ré, C. and Koch, C. Morgan & Claypool, 2011.
“Query Processing on Probabilistic Data: A Survey». Van den Broeck, G. and Suciu, D. Foundations and Trends in Databases Vol. 7, No. 3-4 (2015) 197-341. NOW Publishers.

“Statistical Relational Artificial Intelligence». De Raedt, L., Kersting, K., Natarajan, S. and Poole, D. Morgan & Claypool, 2016.

“Foundations of Probabilistic Logic Programmning». Riguzzi, F. River Publishers, 2018.

“An Introduction to Lifted Probabilistic Inference». Van den Broeck, G., Kersting, K., Natajaran, S. and Poole, D. The MIT Press, 2021.

“Bayesian Networks and Decision Graphs». Jensen, F. V. and Nielsen, T. D.
2nd Ed., Springer 2007.

Previuos requirements

Basic notions on probability theory. Some knowledge of relational databases and classical logic would be expected. However, I would make some personal background material on these two subjects available before the course.

Inscripción (próximamente)
Ver todos los cursos