HeterSEED: Semantics-Structure Decoupling for Heterogeneous Graph Learning under Heterophily

TL;DR

HeterSEED introduces a novel framework for heterogeneous graph learning that decouples semantics and structure, effectively handling heterophily and outperforming existing methods on large-scale real-world graphs.

Contribution

The paper proposes HeterSEED, a semantics-structure decoupling approach that enhances heterophily handling and theoretical expressiveness in heterogeneous graph neural networks.

Findings

HeterSEED outperforms existing GNNs on five real-world graphs.

It is especially effective in strongly heterophilic regimes.

HeterSEED scales to large networks with millions of nodes.

Abstract

Many real-world heterogeneous graphs exhibit pronounced heterophily, where connected nodes often have dissimilar labels or play different semantic roles. In such settings, standard heterogeneous graph neural networks that aggregate messages along metapaths or meta-relations primarily based on feature similarity can propagate misleading information, since feature similarity may be misaligned with underlying relational semantics. In this paper, we propose HeterSEED, a semantics-structure decoupling framework for heterogeneous graph learning under heterophily. HeterSEED decouples representation learning into a heterogeneous semantic channel that captures type- and relation-aware local semantics and a structure-aware heterophily channel that separates homophilic and heterophilic neighborhoods via pseudo-label-guided partitioning and aggregates them using metapath-based structural weights. A node-level adaptive fusion mechanism then combines the two channels to produce context-dependent node representations. Theoretically, we establish that, on heterogeneous graphs under heterophily, HeterSEED is strictly more expressive than standard heterogeneous graph neural networks that rely primarily on feature similarity and provably reduces the prediction bias introduced by heterophilic neighbors. Experiments on five real-world heterogeneous graphs, including two large-scale networks at the million-node and hundred-million-edge scale, demonstrate that HeterSEED consistently outperforms representative heterogeneous graph neural networks and recent heterophily-aware baselines, especially in strongly heterophilic regimes.