| Abstract | Most real-world data is stored in relational form. Incontrast, most statistical learning methods work with
“flat” data representations, forcing us to convert our
data into a form that loses much of the relational struc-
ture. The recently introduced framework of proba-
bilistic relational models (PRMs) allows us to repre-
sent probabilistic models over multiple entities that
utilize the relations between them. In this paper, we
propose the use of probabilistic models not only for
the attributes in a relational model, but for the rela-
tional structure itself. We propose two mechanisms for
modeling structural uncertainty: reference uncertainty
and existence uncertainty. We describe the appropriate
conditions for using each model and present learning
algorithms for each. We present experimental results
showing that the learned models can be used to pre-
dict relational structure and, moreover, the observed
relational structure can be used to provide better pre-
dictions for the attributes in the model.
|
