Probability Logic and Statistical Relational Artificial Intelligence

After decades of study in the philosophy community, probability logic established itself as a knowledge representation formalism for quantifying uncertainty in artificial intelligence in the late 1980s. Since then, however, statistical relational artificial intelligence emerged as a new paradigm for combining first-order reasoning with probabilistic uncertainty, harnessing the rapid advances made in the field of probabilistic graphical models. In this contribution, we suggest how first-order logics of probability, and the taxonomy developed for them around 30 years ago, can contribute to a deeper understanding of statistical relational frameworks, enhance the expressivity of their models and formulate queries of practical significance and theoretical interest.