We are privileged to have the following speakers in our workshop

Bart Selman Cornell University, USA
The Synthesis of Probabilistic and Logical Inference Methods

In recent years, constraint reasoning methods have improved dramatically. Up till the mid nineties, general constraint reasoning beyond hundred variable problems appeared infeasible. Since then, we have witness a qualitative change in the field: current reasoning engines can handle problems with over a million variables and several millions of constraints. I will discuss what led to such a dramatic scale-up, emphasizing recent advances on combining probabilistic and logical inference techniques.

Josh Tenenbaum MIT, USA
The structure, function and acquisition of common-sense theories: a StarAI perspective

Human beings organize and interpret their experience through the lens of large-scale systems of abstract knowledge. These systems resemble scientific theories in certain ways and are often referred to as "intuitive theories" or "folk theories", as in "intuitive physics" or "folk psychology". A long-term goal of cognitive science is to explain how these theories are structured, how they are used and how they are learned. Likewise, a major goal of AI is to capture these theories in computational terms, in order to endow machines with human-like common sense. The tools of statistical and relational AI (StarAI) are uniquely well-suited to this project. I will show how cognitive scientists have been applying well-known StarAI tools, and developing new ones, in the service of reverse-engineer people's intuitive theories. I will also discuss some lessons and challenges for engineering this kind of knowledge in future AI systems.

This talk will cover joint work with Noah Goodman, Charles Kemp, Tomer Ullman, Chris Baker, and Ed Vul.