Knowledge Representation and Reasoning with Nilsson-style Probabilistic Logics
A natural idea to extend classical logics to probabilistic logics is to consider probability distributions over classical logical interpretations. While similar ideas have been considered by Boole already, Nilsson made this approach popular in Artificial Intelligence. Since then, various propositional and (propositionalized) relational logics have been investigated that fall into this framework. The tutorial will give an in-depth discussion of the basic propositional setting (conditional probabilities over formulas), basic reasoning problems and basic and advanced approaches to solve these problems. We will also discuss extensions to deal with inconsistencies and more expressive languages, in particular relational ones.
The tutorial will be mostly self-contained. While I assume some familiarity with basic logic and probability theory, we will revisit all necessary definitions. The tutorial is therefore potentially interesting for all researchers interested in symbolic uncertain reasoning.