Identifying Independence in Relational Models

TL;DR

This paper introduces relational d-separation, extending the concept of conditional independence to relational models, with a sound, complete, and efficient method supported by empirical validation.

Contribution

It develops a novel theory and computational method for deriving conditional independence in relational models, expanding beyond simple directed graphical models.

Findings

Effective relational d-separation method demonstrated

Sound and complete theoretical framework established

Empirical results confirm practical applicability

Abstract

The rules of d-separation provide a framework for deriving conditional independence facts from model structure. However, this theory only applies to simple directed graphical models. We introduce relational d-separation, a theory for deriving conditional independence in relational models. We provide a sound, complete, and computationally efficient method for relational d-separation, and we present empirical results that demonstrate effectiveness.