Using First-Order Probability Logic for the Construction of Bayesian Networks

TL;DR

This paper introduces a method to build Bayesian networks from a rich first-order probabilistic logic knowledge base, enabling more flexible and extensive model construction than traditional template methods.

Contribution

It presents a novel approach that uses first-order probabilistic logic for constructing Bayesian networks from general knowledge bases, expanding the range of possible models.

Findings

Supports construction of diverse Bayesian networks

Utilizes local statistical information for model building

Practical for rich, complex knowledge bases

Abstract

We present a mechanism for constructing graphical models, specifically Bayesian networks, from a knowledge base of general probabilistic information. The unique feature of our approach is that it uses a powerful first-order probabilistic logic for expressing the general knowledge base. This logic allows for the representation of a wide range of logical and probabilistic information. The model construction procedure we propose uses notions from direct inference to identify pieces of local statistical information from the knowledge base that are most appropriate to the particular event we want to reason about. These pieces are composed to generate a joint probability distribution specified as a Bayesian network. Although there are fundamental difficulties in dealing with fully general knowledge, our procedure is practical for quite rich knowledge bases and it supports the construction of a far wider range of networks than allowed for by current template technology.