Text-attributed graphs (TAGs), which associate rich textual descriptions with each node, are widely employed to represent complex relationships among real-world textual entities. Currently, representation learning for TAGs leverages large language models (LLMs) to transform node-matched textual descriptions into node features or labels, followed by the message passing in graph neural networks (GNNs) that further improves the expressiveness of graph representation learning. Nevertheless, a simple experiment we conducted demonstrates that not all LLMs are readily compatible with GNNs. A salient finding indicates that architectural heterogeneity among LLMs manifests as substantial performance gap across diverse TAGs representation learning. Moreover, the node semantics encoded by LLMs are often misaligned with the message passing in GNNs, causing performance collapse. Motivated by this observation, we propose a novel self-supervised graph learning framework called Stage-Aware Graph Contrastive Learning (SAGCL). In particular, we propose the node-oriented mixture of experts (NodeMoE) to assign suitable candidate experts for each node. It flexibly balances the strengths of different language experts by low-rank decomposition and reparameterization strategies. Subsequently, to align the inductive biases of graph structures with the semantic perception capabilities of LLMs, the message passing in GNNs is decoupled into the feature transformation stage and the feature propagation stage. Given the two stage views, stage-aware graph contrastive learning is proposed to match the node semantics encoded by the LLM with the locally aware topological patterns within the GNN via self-supervised contrastive learning. Experiments on eight datasets and three downstream tasks demonstrate the effectiveness of SAGCL.

Dynamic graph neural networks (DyGNNs) currently struggle with handling distribution shifts that naturally arise when training and test data follow similar but non-identical distributions. As the generation of dynamic graphs is strongly influenced by latent environments, it is critical to investigate their impacts on the generalization behavior of DyGNNs. We therefore establish a connection between the temporal message-passing scheme employed by DyGNNs and their generalization performance under distribution shifts. Our analysis reveals that environment-specific factors misguide the learning process and lead to unsatisfactory out-of-distribution (OOD) generalization. Based on this insight, we propose MoVE, a Mixture of Variational Experts network to mitigate complex distribution shifts in dynamic graphs. MoVE adopts a hierarchical variational architecture that extrapolates latent representations into a mixture of distribution shifts as pseudo-environments. Additionally, we incorporate a Mixture-of-Experts (MoE) framework with a novel training objective that aligns the outputs of different experts to produce invariant representations. Extensive experiments on various dynamic graphs, including both real-world and synthetic datasets, demonstrate that our model significantly outperforms state-of-the-art techniques.

Federated graph learning (FGL), as a privacy-preserving paradigm for distributed graph data training, aims to resolve graph data isolation issues under the framework of federated learning (FL). Despite the significant efforts made by existing FGL methods, two key challenges are still not well addressed: 1) how to mitigate graph heterogeneity in clients arising from feature deviation and structural deviation and 2) how to devise a favorable aggregation mechanism to maximize the client's benefit from collaborative training with privacy preserving. To tackle these issues, we take a perspective of latent factor and propose a HarmoFGL framework by Harmonizing graph neural network (GNN) latent factors for Federated Graph Learning, achieving cross-client federated training by coordinating personalized aggregation and client-level representation in a symbiotic space. To alleviate feature deviation, an implicit feature crossing (IFC) approach is proposed through the disentanglement of higher order feature dependency into client-universal and client-specific interactions. As for the graph heterogeneity induced by structural deviation, we establish a cross-client symbiotic parameter space spanned by GNN latent factors, on which a client-level representation is derived to characterize the inherent properties of clients. On the server side, on the basis of client relevance-driven personalized parameter aggregation, graph Laplacian regularization on client-level representations is implemented for collaborative training. Experimental results on five public graph datasets and two medical datasets demonstrate the effectiveness of HarmoFGL.

Recently, post-training quantization (PTQ) methods for large language models (LLMs) primarily focus on tackling the challenges caused by outliers. Scaling transformation has proven to be effective while how to enhance the performance of extremely low-bitwidth (e.g., 2-bit) PTQ under it remains largely unexplored. In this work, a new PTQ framework, namely MPPQ, is established. Specifically, MPPQ first proposes an enhanced reconstruction loss based on Mixed metric supervision to mitigate the distribution inconsistency caused by quantization while providing strong regularization for learnable parameters. Secondly, we introduce a Proxy-based adaptive rounding scheme in weight quantization, which replaces the round-to-nearest (RTN) function to minimize the overall quantization errors through element-wise scaling. Furthermore, a factor coarse Pre-searching mechanism is presented to ensure proper coordination between quantization and clipping patterns, while achieving optimal initialization of clipping factors before training. Extensive experiments show that MPPQ consistently outperforms state-of-the-art methods in low-bit quantization settings. For instance, the perplexity of WikiText2 can be dramatically reduced to 8.85 (3.9 ↓ vs 12.75 of the latest method, LRQuant) for the LLaMA-2-7B model, which is quantized with W4A4.

Graph condensation (GC) aims to distill the original graph into a small-scale graph, mitigating redundancy and accelerating GNN training. However, conventional GC approaches heavily rely on rigid GNNs and task-specific supervision. Such a dependency severely restricts their reusability and generalization across various tasks and architectures. In this work, we revisit the goal of ideal GC from the perspective of GNN optimization consistency, and then a generalized GC optimization objective is derived, by which those traditional GC methods can be viewed nicely as special cases of this optimization paradigm. Based on this, \textbf{Pre}-trained \textbf{G}raph \textbf{C}ondensation (\textbf{PreGC}) via optimal transport is proposed to transcend the limitations of task- and architecture-dependent GC methods. Specifically, a hybrid-interval graph diffusion augmentation is presented to suppress the weak generalization ability of the condensed graph on particular architectures by enhancing the uncertainty of node states. Meanwhile, the matching between optimal graph transport plan and representation transport plan is tactfully established to maintain semantic consistencies across source graph and condensed graph spaces, thereby freeing graph condensation from task dependencies. To further facilitate the adaptation of condensed graphs to various downstream tasks, a traceable semantic harmonizer from source nodes to condensed nodes is proposed to bridge semantic associations through the optimized representation transport plan in pre-training. Extensive experiments verify the superiority and versatility of PreGC, demonstrating its task-independent nature and seamless compatibility with arbitrary GNNs.

Federated graph learning (FGL) is a promising distributed training paradigm for graph neural networks across multiple local systems without direct data sharing. This approach inherently involves large-scale distributed graph processing, which closely aligns with the challenges and research focuses of graph-based data systems. Despite the proliferation of FGL, the diverse motivations from real-world applications, spanning various research backgrounds and settings, pose a significant challenge to fair evaluation. To fill this gap, we propose OpenFGL, a unified benchmark designed for the primary FGL scenarios: Graph-FL and Subgraph-FL. Specifically, OpenFGL includes 42 graph datasets from 18 application domains, 8 federated data simulation strategies that emphasize different graph properties, and 5 graph-based downstream tasks. Additionally, it offers 18 recently proposed SOTA FGL algorithms through a user-friendly API, enabling a thorough comparison and comprehensive evaluation of their effectiveness, robustness, and efficiency. Our empirical results demonstrate the capabilities of FGL while also highlighting its potential limitations, providing valuable insights for future research in this growing field, particularly in fostering greater interdisciplinary collaboration between FGL and data systems.

Due to the widespread applications of heterogeneous graphs in the real world, heterogeneous graph neural networks (HGNNs) have developed rapidly and made a great success in recent years. To effectively capture the complex interactions in heterogeneous graphs, various attention mechanisms are widely used in designing HGNNs. However, the employment of these attention mechanisms brings two key problems: high computational complexity and poor robustness. To address these problems, we propose aFastandRobust attention-freeHeterogeneousGraphConvolutionalNetwork (FastRo-HGCN) without any attention mechanisms. Specifically, we first construct virtual links based on the topology similarity and feature similarity of the nodes to strengthen the connections between the target nodes. Then, we design type normalization to aggregate and transfer the intra-type and inter-type node information. The above methods are used to reduce the interference of noisy information. Finally, we further enhance the robustness and relieve the negative effects of oversmoothing with the self-loops of nodes. Extensive experimental results on three real-world datasets fully demonstrate that the proposed FastRo-HGCN significantly outperforms the state-of-the-art models. The codes and data of this work are available athttps://github.com/ZZY-GraphMiningLab/FastRo-HGCN.