HSDP: Hypergraph and structure-aware representation learning for information diffusion prediction

doi:10.1016/j.patcog.2026.113732

HSDP: Hypergraph and structure-aware representation learning for information diffusion prediction

2026-04-15

0

PRE

AI

DOI：10.1016/j.patcog.2026.113732

Original

Original request for help

Save

Abstract

En 中文

• We propose a hypergraph structure-aware framework that unifies dual graph auto-encoding and hypergraph modeling for information diffusion prediction. • The model captures both structural heterogeneity and inter-cascade dependencies, with gated fusion and sequential modeling to adaptively balance multiple representations. • Experiments on four real-world datasets demonstrate that HSDP consistently outperforms state-of-the-art baselines, validated by ablation and hyperparameter analyses.

Keywords:

hypergraph modeling

information diffusion prediction

dual graph auto-encoding

structural heterogeneity

gated fusion

HSDP: Hypergraph and structure-aware representation learning for information diffusion prediction

Abstract

Journal

Organization