Graph Clustering with Cross-View Feature Propagation

TL;DR

GCCFP addresses graph clustering by jointly leveraging graph topology and multi-view vertex features through cross-view feature propagation. It introduces a unified objective $O$ that combines structural propagation and cross-view feature reconstruction and solves it with a convergent block-coordinate algorithm updating $V$, $U$, $X$, $P^i$, $C$, and $W$. The approach regularizes learning with latent propagation factors and enforces nonnegativity and partial orthogonality to capture pivotal information. Empirical evaluation on ten real-world graphs shows GCCFP achieving state-of-the-art clustering performance across diverse domains, demonstrating robustness to varying view counts and feature types.

Abstract

Graph clustering is a fundamental and challenging learning task, which is conventionally approached by grouping similar vertices based on edge structure and feature similarity.In contrast to previous methods, in this paper, we investigate how multi-view feature propagation can influence cluster discovery in graph data.To this end, we present Graph Clustering With Cross-View Feature Propagation (GCCFP), a novel method that leverages multi-view feature propagation to enhance cluster identification in graph data.GCCFP employs a unified objective function that utilizes graph topology and multi-view vertex features to determine vertex cluster membership, regularized by a module that supports key latent feature propagation. We derive an iterative algorithm to optimize this function, prove model convergence within a finite number of iterations, and analyze its computational complexity. Our experiments on various real-world graphs demonstrate the superior clustering performance of GCCFP compared to well-established methods, manifesting its effectiveness across different scenarios.

Figures (3)

• Figure 1: Overall idea of the proposed Graph Clustering With Cross-View Feature Propagation (GCCFP).
• Figure 2: Sensitivity test of hyper-parameters $\alpha$ and $\lambda$ on $Ego-facebook$.
• Figure 3: Model convergence on testing datasets.