HeteGraph-Mamba: Heterogeneous Graph Learning via Selective State Space Model

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Pan, Zhenyu, Jeong, Yoonsung, Liu, Xiaoda, Liu, Han
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917673768583168
author Pan, Zhenyu
Jeong, Yoonsung
Liu, Xiaoda
Liu, Han
author_facet Pan, Zhenyu
Jeong, Yoonsung
Liu, Xiaoda
Liu, Han
contents We propose a heterogeneous graph mamba network (HGMN) as the first exploration in leveraging the selective state space models (SSSMs) for heterogeneous graph learning. Compared with the literature, our HGMN overcomes two major challenges: (i) capturing long-range dependencies among heterogeneous nodes and (ii) adapting SSSMs to heterogeneous graph data. Our key contribution is a general graph architecture that can solve heterogeneous nodes in real-world scenarios, followed an efficient flow. Methodologically, we introduce a two-level efficient tokenization approach that first captures long-range dependencies within identical node types, and subsequently across all node types. Empirically, we conduct comparisons between our framework and 19 state-of-the-art methods on the heterogeneous benchmarks. The extensive comparisons demonstrate that our framework outperforms other methods in both the accuracy and efficiency dimensions.
format Preprint
id arxiv_https___arxiv_org_abs_2405_13915
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle HeteGraph-Mamba: Heterogeneous Graph Learning via Selective State Space Model
Pan, Zhenyu
Jeong, Yoonsung
Liu, Xiaoda
Liu, Han
Machine Learning
Social and Information Networks
We propose a heterogeneous graph mamba network (HGMN) as the first exploration in leveraging the selective state space models (SSSMs) for heterogeneous graph learning. Compared with the literature, our HGMN overcomes two major challenges: (i) capturing long-range dependencies among heterogeneous nodes and (ii) adapting SSSMs to heterogeneous graph data. Our key contribution is a general graph architecture that can solve heterogeneous nodes in real-world scenarios, followed an efficient flow. Methodologically, we introduce a two-level efficient tokenization approach that first captures long-range dependencies within identical node types, and subsequently across all node types. Empirically, we conduct comparisons between our framework and 19 state-of-the-art methods on the heterogeneous benchmarks. The extensive comparisons demonstrate that our framework outperforms other methods in both the accuracy and efficiency dimensions.
title HeteGraph-Mamba: Heterogeneous Graph Learning via Selective State Space Model
topic Machine Learning
Social and Information Networks
url https://arxiv.org/abs/2405.13915