# Learning Robust Heterogeneous Graph Representations via Contrastive-Reconstruction under Sparse Semantics

arXiv: https://arxiv.org/abs/2506.06682 · HTML: https://mlsanity.com/paper/2506.06682/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 35 references that are in the map.
- First posted on arXiv 7 Jun 2025.

## Builds on (most cited first, 19 of 19)

- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Self-supervised Learning: Generative or Contrastive](https://mlsanity.com/paper/2006.08218/) (2020)
- [Graph Contrastive Learning with Adaptive Augmentation](https://mlsanity.com/paper/2010.14945/) (2020)
- [GraphMAE: Self-Supervised Masked Graph Autoencoders](https://mlsanity.com/paper/2205.10803/) (2022)
- [MAGNN: Metapath Aggregated Graph Neural Network for Heterogeneous Graph Embedding](https://mlsanity.com/paper/2002.01680/) (2020)
- [Graph Neural Networks for Graphs with Heterophily: A Survey](https://mlsanity.com/paper/2202.07082/) (2022)
- [Self-supervised Heterogeneous Graph Neural Network with Co-contrastive Learning](https://mlsanity.com/paper/2105.09111/) (2021)
- [When Does Self-Supervision Help Graph Convolutional Networks?](https://mlsanity.com/paper/2006.09136/) (2020)
- [Rethinking Graph Masked Autoencoders through Alignment and Uniformity](https://mlsanity.com/paper/2402.07225/) (2024)
- [Graph Contrastive Learning with Personalized Augmentation](https://mlsanity.com/paper/2209.06560/) (2022)
- [Relation-aware Graph Attention Model With Adaptive Self-adversarial Training](https://mlsanity.com/paper/2102.07186/) (2021)
- [Learning Combinatorial Node Labeling Algorithms](https://mlsanity.com/paper/2106.03594/) (2021)
- [Heterogeneous Graph Contrastive Learning with Meta-path Contexts and Adaptively Weighted Negative Samples](https://mlsanity.com/paper/2212.13847/) (2022)
- [DiffGraph: Heterogeneous Graph Diffusion Model](https://mlsanity.com/paper/2501.02313/) (2025)
- [Hop-Hop Relation-aware Graph Neural Networks](https://mlsanity.com/paper/2012.11147/) (2020)
- [Positive-Negative Equal Contrastive Loss for Semantic Segmentation](https://mlsanity.com/paper/2207.01417/) (2022)
- [LocalGCL: Local-aware Contrastive Learning for Graphs](https://mlsanity.com/paper/2402.17345/) (2024)
- [NodeReg: Mitigating the Imbalance and Distribution Shift Effects in Semi-Supervised Node Classification via Norm Consistency](https://mlsanity.com/paper/2503.03211/) (2025)

## Abstract (arXiv)

In graph self-supervised learning, masked autoencoders (MAE) and contrastive learning (CL) are two prominent paradigms. MAE focuses on reconstructing masked elements, while CL maximizes similarity between augmented graph views. Recent studies highlight their complementarity: MAE excels at local feature capture, and CL at global information extraction. Hybrid frameworks for homogeneous graphs have been proposed, but face challenges in designing shared encoders to meet the semantic requirements of both tasks. In semantically sparse scenarios, CL struggles with view construction, and gradient imbalance between positive and negative samples persists. This paper introduces HetCRF, a novel dual-channel self-supervised learning framework for heterogeneous graphs. HetCRF uses a two-stage aggregation strategy to adapt embedding semantics, making it compatible with both MAE and CL. To address semantic sparsity, it enhances encoder output for view construction instead of relying on raw features, improving efficiency. Two positive sample augmentation strategies are also proposed to balance gradient contributions. Node classification experiments on four real-world heterogeneous graph datasets demonstrate that HetCRF outperforms state-of-the-art baselines. On datasets with missing node features, such as Aminer and Freebase, at a 40% label rate in node classification, HetCRF improves the Macro-F1 score by 2.75% and 2.2% respectively compared to the second-best baseline, validating its effectiveness and superiority.