Knowledge Propagation over Conditional Independence Graphs

TL;DR

This paper introduces algorithms for knowledge propagation over Conditional Independence graphs, enhancing understanding of feature dependencies and outperforming existing methods on benchmark datasets.

Contribution

It presents novel algorithms for propagating knowledge in CI graphs, a new approach in modeling feature dependencies in probabilistic graphical models.

Findings

Improved performance over state-of-the-art on Cora dataset

Enhanced insights into domain topology from CI graphs

Effective knowledge propagation algorithms demonstrated

Abstract

Conditional Independence (CI) graph is a special type of a Probabilistic Graphical Model (PGM) where the feature connections are modeled using an undirected graph and the edge weights show the partial correlation strength between the features. Since the CI graphs capture direct dependence between features, they have been garnering increasing interest within the research community for gaining insights into the systems from various domains, in particular discovering the domain topology. In this work, we propose algorithms for performing knowledge propagation over the CI graphs. Our experiments demonstrate that our techniques improve upon the state-of-the-art on the publicly available Cora and PubMed datasets.