mlsanityData to 14 Oct 2026

THGFM: Dual-Branch Temporal Heterogeneous Graph Fusion Model

arXiv 2607.27303 ↗First posted 29 Jul 2026cs.LG cs.CL stat.MLcited by 0

Not on the map. Why?

Abstract (arXiv)

Temporal heterogeneous graphs offer a natural abstraction for dynamic relational systems in which diverse node and relation types co-exist and evolve over time. Learning on such graphs requires jointly modeling cross-type structural heterogeneity and the temporal dynamics of interactions, yet existing methods still struggle to reconcile parameter-efficient cross-type transfer with relation-aware specialization, and typically inject time only as additive features outside the attention kernel. We propose THGFM, a web-scale temporal heterogeneous graph fusion model that addresses both limitations within a unified dual-path architecture. THGFM couples a textitShared-Space Temporal Attention branch for parameter-efficient cross-type transfer with a textitRelational Type-Partitioned Temporal Attention branch for relation-aware specialization, and integrates them through textitDual-Path Relational--Shared Fusion, instantiated with textitType-Conditioned Non-Competitive Gated Sum Fusion: a adaptive mechanism that assigns independent, type-conditioned feature-wise gates to the shared and specialized branches, allowing both to be amplified or suppressed without zero-sum competition. To directly incorporate relative time into the attention score, THGFM further introduces textitRotary Temporal Attention, which rotates queries and keys by half-phases of relative time before matching. THGFM consistently outperforms baseline graph transformer models on academic graphs benchmarks, delivering a +3.25% six-task mean gain, with peak relative gains of +12.37% on OAG-CS PV, +4.87% on PF-L2, and +1.18% on PF-L1, and +4.24%, +3.73%, and +4.61% on OGBN-MAG, HTAG-ArXiv, and HTAG-DBLP, respectively.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 22 of its 32 references that are in the map. First posted on arXiv 29 Jul 2026.

Builds on

22 of 32 references.

References · 22

  1. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
  2. 2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,811
  3. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,163
  4. 2017Modeling Relational Data with Graph Convolutional Networkscited by 1,694
  5. 2019Heterogeneous Graph Attention Networkcited by 734
  6. 2020Knowledge Graphscited by 621
  7. 2020Heterogeneous Graph Transformercited by 582
  8. 2020Temporal Graph Networks for Deep Learning on Dynamic Graphscited by 530
  9. 2020Inductive Representation Learning on Temporal Graphscited by 462
  10. 2019Composition-based Multi-Relational Graph Convolutional Networkscited by 396
  11. 2020MAGNN: Metapath Aggregated Graph Neural Network for Heterogeneous Graph Embeddingcited by 304
  12. 2021Are we really making much progress? Revisiting, benchmarking, and refining heterogeneous graph neural networkscited by 176
10 more
  1. 2019An Attention-based Graph Neural Network for Heterogeneous Structural Learningcited by 71
  2. 2022Simple and Efficient Heterogeneous Graph Neural Networkcited by 71
  3. 2020Interpretable and Efficient Heterogeneous Graph Convolutional Networkcited by 48
  4. 2021Heterogeneous Graph Representation Learning with Relation Awarenesscited by 31
  5. 2024A Comprehensive Survey of Dynamic Graph Neural Networks: Models, Frameworks, Benchmarks, Experiments and Challengescited by 30
  6. 2021Heterogeneous Temporal Graph Neural Networkcited by 23
  7. 2024Multi-Scale Heterogeneous Text-Attributed Graph Datasets From Diverse Domainscited by 5
  8. 2026Billion-Scale Graph Foundation Modelscited by 5
  9. 2024CTRL: Continuous-Time Representation Learning on Temporal Heterogeneous Information Networkcited by 2
  10. 2024Heta: Distributed Training of Heterogeneous Graph Neural Networkscited by 2

Built on by

Not cited yet.