Bayesian networks provide an elegant formalism for representing and
reasoning about uncertainty using probability theory. Stuart Russell once
stated that they are "the best thing since sliced bread". However,
Bayesian networks are a probabilistic extension of propositional logic
and, hence, inherit some of the limitations of propositional logic, such
as the difficulties to represent objects and relations. Consider e.g.
to state that "the person X's height is influenced by the heights of its
mother M and its father F".

This talk introduces a generalization of Bayesian networks, 
called Bayesian logic programs, to overcome these limitations. 
In order to represent objects and relations it combines Bayesian networks 
with definite clause logic by establishing a one-to-one mapping 
between ground atoms and random variables. It will be shown that Bayesian
logic programs combine the advantages of both definite clause logic and 
Bayesian networks. This includes the separation of quantitative and
qualitative aspects of the model. Furthermore, Bayesian logic programs 
generalize both Bayesian networks as well as logic programs. So, many
ideas developed in both areas carry over.

To illustrate Bayesian logic programs, an example from the field 
of genetics will be employed. Genetics provide an intuitive and natural 
application domain for first order probabilistic models, because it has a 
probabilistic nature given by the biological laws of inheritance, and
requires the representation of the relational familial structure of the
individuals under study. A demo of an interpreter for Bayesian logic
programs will complete the talk.