Spectral clustering of annotated graphs using a factor graph representation

TL;DR

This paper introduces a spectral clustering method for annotated graphs that encodes structural and annotation information as a factor graph, providing a new mathematical foundation for this approach.

Contribution

It develops a novel spectral clustering framework based on factor graph representations of annotated graphs, bridging graph structure and annotations mathematically.

Findings

Provides a mathematical basis for spectral clustering with factor graphs

Demonstrates the effectiveness of the method on annotated graph data

Offers a new perspective on integrating annotations into graph clustering

Abstract

Graph-structured data commonly have node annotations. A popular approach for inference and learning involving annotated graphs is to incorporate annotations into a statistical model or algorithm. By contrast, we consider a more direct method named scotch-taping, in which the structural information in a graph and its node annotations are encoded as a factor graph. Specifically, we establish the mathematical basis of this method in the spectral framework.