Efficient and Scalable Multiview Masked Autoencoder for Heterogeneous Graph Self-Supervised Learning

Recent years have witnessed the flourishing of self-supervised learning (SSL) on heterogeneous graphs, e.g., masked autoencoder (MAE)-based methods. However, existing MAE-based methods mostly suffer from three issues, including the prohibitive computational cost of constructing meta-path-induced subgraphs, the limitations of uniform random masking, and the suboptimal objective of reconstructing potentially noisy or uninformative original node features. To address these issues, we propose the multiview masked autoencoder (MVMAE)-a novel, efficient, and scalable framework for heterogeneous graph SSL. First, MVMAE introduces a novel multiview mask-then-reconstruct strategy. Instead of masking individual nodes or features, MVMAE randomly masks entire meta-path views and guides the model to reconstruct their representations from remaining visible views. This approach shifts the learning objective to a high-level, semantic task, providing a more robust and informative supervisory signal that is effective even when initial node features are absent or flawed. Then, to handle the computational demands, MVMAE employs an efficient offline aggregation mechanism that precomputes and reuses representations for each meta-path view, completely avoiding the costly construction of meta-path-induced subgraphs. Next, to fuse the representations of each view, MVMAE uses a Transformer-based multiview fusion encoder to capture the complex correlations between views, and a deliberately simple MLP decoder to compel the encoder to learn highly expressive and robust node representations. Extensive experiments on six real-world datasets, including three large-scale graphs, demonstrate that MVMAE consistently outperforms state-of-the-art self-supervised and even semisupervised baselines, confirming its superior effectiveness and scalability. Our source code is released at https://github.com/hcmdgh/MVMAE.