Out-of-Distribution (OOD) generalization is a promising yet challenging goal that guarantees the test performance of Graph Neural Networks (GNNs) in open-world settings. However, due to the intricate internal topology of graph-structured data, redundant information from the spurious topologies severely confuses GNNs to deviate from the labels. Extracting concise and label-relevant subgraphs from the original graphs can alleviate this problem. Unfortunately, existing methods either overlook the global structural distribution or rely heavily on manually predefined assumptions. As a result, they fall short of well capturing the structural distribution changes between input graph and extracted subgraph, thus compromising adaptability of extracted invariant subgraphs to diverse OOD scenarios. This motivates us to propose a framework called S tructural- E ntropy-guided I nformation B ottleneck (OOD-SEIB) that aims to more traceably measure the inherent information changes for better and more flexible OOD generalization. The core of OOD-SEIB lies in concise topology extraction module, where we measure the mutual information flow between input graph and extracted subgraph based on structural entropy, termed Compression Index (CI). Specifically, the CI is a quantifiable metric that calculates the codeword length required to describe entire graph structure via a biased random walk. Under this guidance, OOD-SEIB then launches a structural information bottleneck compression module that jointly optimizes both CI and label-relevance of the subgraph topology by iteratively balancing between informativeness and compression. To further improve GNN’s invariant subgraph identification capability, OOD-SEIB generates multiple augmented environments and distill the invariant subgraphs into GNN as knowledge in an inside-out manner. When iteratively optimizing in above prescribed way, OOD-SEIB progressively reinforce the invariant subgraph extraction, thereby enhancing its generalization capability. Extensive experiments on synthetic and three real-world graph-level OOD benchmarks demonstrate that our proposed OOD-SEIB improves classification accuracy by 4.85%–38.03% on average compared to state-of-the-art baselines. Additionally, we extend OOD-SEIB to two node-level benchmarks, achieving average classification accuracy improvements of 14.52% and 13.15%.

The deployment of sensors enables data-driven urban management, but necessitates inductive spatio-temporal kriging to infer unmonitored areas. Existing methods impute these unknown observations by smoothing temporal features based on spatial dependencies, overlooking the decoupling ofinherent propertiesanddynamic correlationsin message passing. In particular, the inherent properties reveal non-transitive signals, and current coupled aggregation leads to inaccurate results. To this end, we proposeTempoRAl deCoupledKriging, named TRACK, to decouple two factors with the help of node-specific inherency. Specifically, we first construct a node-specific profile to represent its inherency including geographical and periodic features, which is subsequently transformed into decoupling prompts. Secondly, the coupled temporal features are separated through querying each prompt embedding, facilitating precise temporal aggregation for inherent properties and spatial aggregation for dynamic correlations. Finally, a multi-task training strategy is further adopted to mimic the inductive scenarios during testing. We evaluate TRACK on four real-world datasets spanning urban traffic and air quality prediction tasks. TRACK achieves state-of-the-art performance, with average improvements of 3.10% in MAE and 4.45% in RMSE over strong baselines. Moreover, we further demonstrated its robust generalization in a challenging cross-city inductive setting. Code is available athttps://github.com/JeremyChou28/TRACK.

Reasoning over Knowledge Graphs (KGs) plays a pivotal role in knowledge graph completion or question answering systems, providing richer and more accurate triples and attributes. As numerical attributes become increasingly essential in characterizing entities and relations in KGs, the ability to reason over these attributes has gained significant importance. Existing graph-based methods such as Graph Neural Networks (GNNs) and Knowledge Graph Embeddings (KGEs), primarily focus on aggregating homogeneous local neighbors and implicitly embedding diverse triples. However, these approaches often fail to fully leverage the potential of logical paths within the graph, limiting their effectiveness in exploiting the reasoning process. To address these limitations, we propose ChainsFormer, a novel chain-based framework designed to support numerical reasoning. Chainsformer not only explicitly constructs logical chains but also expands the reasoning depth to multiple hops. Specially, we introduces Relation-Attribute Chains (RA-Chains), a specialized logic chain, to model sequential reasoning patterns. ChainsFormer captures the step-by-step nature of multi-hop reasoning along RA-Chains by employing sequential in-context learning. To mitigate the impact of noisy chains, we propose a hyperbolic affinity scoring mechanism that selects relevant logic chains in a variable-resolution space. Furthermore, ChainsFormer incorporates an attention-based numerical reasoner to identify critical reasoning paths, enhancing both reasoning accuracy and transparency. Experimental results demonstrate that ChainsFormer significantly outperforms state-of-the-art methods, achieving up to a 20.0% improvement in performance. The implementations are available at https://github.com/zhaodazhuang2333/ChainsFormer.

Evolving graphs are prevalent in real-world systems where nodes and edges continuously expand over time, resulting in persistent topological changes across temporal snapshots. As the graph expands, the increasing structural discrepancy between earlier and later timesteps often leads to poor generalization-a phenomenon widely recognized as Temporal out-of-distribution (TOOD) generalization. This issue has spurred growing research interest in generalized graph learning. Existing approaches often rely on modeling latent environments or incorporating structural causal reasoning, but they frequently face scalability limitations when applied to rapidly expanding graphs. In this work, we carry out a detailed analysis of temporal drift in evolving graphs and identify two complementary forms: structural role drift reflected by changes in degree proportions, and semantic drift measured by embedding divergence. To capture this graph evolution, we propose PromptEG, a scalable prompt-based framework that generates node-specific prompts using degree signals and KL-divergence between representations. These prompts are optimized via consistency and contrastive objectives, and injected into GNNs to enhance temporal generalization. Experiments on largescale evolving graphs-containing up to 500 K edges and tens of thousands of nodes-demonstrate that PromptEG achieves strong long-horizon generalization while maintaining high computational efficiency.

Sequential recommendation (SR) is widely deployed in e-commerce platforms, streaming services, etc., revealing significant potential to enhance user experience. The core of SR lies in exploring the sequential relationships in historical user-item interactions. However, existing methods often overlook two critical factors: irregular user interests between interactions and highly uneven item distributions over time. The former factor implies that actual user preferences are not always continuous, and long-term historical interactions may not be relevant to current purchasing behavior. Therefore, relying only on these historical interactions for recommendations may result in a lack of user interest at the target time. The latter factor, characterized by peaks and valleys in interaction frequency, may result from seasonal trends, special events, or promotions. These externally driven distributions may not align with individual user interests, leading to inaccurate recommendations. To address these deficiencies, we propose TGODE to both enhance and capture the long-term historical interactions. Specifically, we first construct the user time graph and item evolution graph, which utilize user personalized preferences and global item distribution information, respectively. To tackle the temporal sparsity caused by irregular user interactions, we design a time-guided diffusion generator to automatically obtain an augmented time-aware user graph. Additionally, we devise a user interest truncation factor to efficiently identify sparse time intervals and achieve balanced preference inference. After that, the augmented user graph and item graph are fed into a generalized graph neural ordinary differential equation (ODE) to align with the evolution of user preferences and item distributions. This allows two patterns of information evolution to be matched over time. Experimental results demonstrate that TGODE outperforms baseline methods across five datasets, with improvements ranging from 10% to 46%. The code is available at https://github.com/Qin-lab-code/TGODE.

With the explosive growth of online multimodal content, multimodal recommender systems(MRSs) have brought significant benefits to multimedia platforms. As MRSs evolve, many studies incorporate advanced technologies like graph neural networks(GNNs) and self-supervised learning(SSL), achieving remarkable results. However, these efforts still suffer from the quality disparity problem. It refers to the mixture of high and low quality across items' multiple modalities, owing to disparities in construction costs or design levels. These low-quality modalities often lack crucial details or introduce noise to the depiction of item, leading to insufficient or polluted item representation. Therefore, we propose a novel framework R2MR: Review and Rewrite Modality for Recommendation to tackle this issue. Specifically, R2MR is composed of two key components: Modality Reviewer and Modality Rewriter. The Modality Reviewer introduces a Consensus Review Mechanism. It performs perspective decomposition based on user representations and learns the consensus quality scores for modalities from diverse perspectives of multiple users. The Modality Rewriter proposes a Latent Mapping Model, which improves the quality of inferior modalities by learning various mapping patterns from high-quality modalities. Comprehensive experiments across three benchmark datasets reveal that R2MR substantially outperforms state-of-the-art methods, achieving an average improvement of 9.20%. The implementations are available at https://github.com/gutang-97/R2MR.

Due to detector malfunctions and communication failures, missing data is ubiquitous during the collection of traffic data. Therefore, it is of vital importance to impute the missing values to facilitate data analysis and decision-making for Intelligent Transportation System (ITS) . However, existing imputation methods generally perform zero pre-filling techniques to initialize missing values, introducing inevitable noise. Moreover, we observe prevalent over-smoothed interpolations, falling short in revealing the intrinsic spatio-temporal correlations of incomplete traffic data. To this end, we propose Mask-Aware Graph Imputation Network (MagiNet) . Our method designs an adaptive mask spatio-temporal encoder to learn the latent representations of incomplete data, eliminating the reliance on pre-filling missing values. Furthermore, we devise a spatio-temporal decoder that stacks multiple blocks to capture the inherent spatial and temporal dependencies within incomplete traffic data, alleviating over-smoothed imputation. Extensive experiments demonstrate that our method outperforms state-of-the-art imputation methods on five real-world traffic datasets, yielding an average improvement of 4.31% in RMSE and 3.72% in MAPE under Missing Completely at Random (MCAR) pattern. Code is available at https://github.com/JeremyChou28/MagiNet .

With the enrichment of user-item interactions, Graph Neural Networks (GNNs) are widely used in recommender systems to alleviate information overload. Nevertheless, they still suffer from the cold-start issue. Knowledge Graphs (KGs), providing external information, have been extensively applied in GNN-based methods to mitigate this issue. However, current KG-aware recommendation methods suffer from the knowledge imbalance problem caused by incompleteness of existing KGs. This imbalance is reflected by the long-tail phenomenon of item attributes, i.e., unpopular items usually lack more attributes compared to popular items. To tackle this problem, we propose a novel framework called EditKG: Editing Knowledge Graph for Recommendation, to balance attribute distribution of items via editing KGs. EditKG consists of two key designs: Knowledge Generator and Knowledge Deleter. Knowledge Generator generates attributes for items by exploring their mutual information correlations and semantic correlations. Knowledge Deleter removes the task-irrelevant item attributes according to the parameterized task relevance score, while dropping the spurious item attributes through aligning the attribute scores. Extensive experiments on three benchmark datasets demonstrate that EditKG significantly outperforms state-of-the-art methods, and achieves 8.98% average improvement. The implementations are available at https://github.com/gutang-97/2024SIGIR-EditKG.

Graph Neural Network (GNN) has demonstrated extraordinary performance in classifying graph properties. However, due to the selection bias of training and testing data (e.g., training on small graphs and testing on large graphs, or training on dense graphs and testing on sparse graphs), distribution deviation is widespread. More importantly, we often observehybrid structure distribution shiftof both scale and density, despite of one-sided biased data partition. The spurious correlations over hybrid distribution deviation degrade the performance of previous GNN methods and show large instability among different datasets. To alleviate this problem, we proposeOOD-GMixupto jointly manipulate the training distribution withcontrollable data augmentationin metric space. Specifically, we first extract the graph rationales to eliminate the spurious correlations due to irrelevant information. Secondly, we generate virtual samples with perturbation on graph rationale representation domain to obtain potential OOD training samples. Finally, we propose OOD calibration to measure the distribution deviation of virtual samples by leveraging Extreme Value Theory, and further actively control the training distribution by emphasizing the impact of virtual OOD samples. Extensive studies on several real-world datasets on graph classification demonstrate the superiority of our proposed method over state-of-the-art baselines.

Graph Neural Networks (GNNs) have achieved great success in semi-supervised node classification. These methods usually assume the closed-set setting and classify unlabeled nodes to known classes. However, in reality, there exits some unknown classes due to biased-sampling, distribution shifts, anomaly, etc. Therefore, it is important to identify unknown classes while classifying known classes, which is defined as Graph Open-Set Recognition (GOSR). To alleviate this problem, we propose Entropy Message Passing (EMP) for GOSR, which takes into account the graph structure information when identifying unknown classes and automatically determines the discrimination threshold. To be specific, we calculate the likelihood of a node belonging to unknown class through entropy propagation. Then, we transform the threshold selection into entropy clustering to identify the unknown class nodes. Finally, we classify the remaining nodes. Experimental evaluations on six benchmark graph datasets demonstrate that our method outperforms state-of-the-art baseline methods in unknown class detection and graph open-set recognition tasks. Especially in unknown class detection, our method has achieved a significant reduction in FPR@95 ranging from 12.62% to 55.88%.

Spatio-temporal graph learning is a key method for urban computing tasks, such as traffic flow, taxi demand and air quality forecasting. Due to the high cost of data collection, some developing cities have few available data, which makes it infeasible to train a well-performed model. To address this challenge, cross-city knowledge transfer has shown its promise, where the model learned from data-sufficient cities is leveraged to benefit the learning process of data-scarce cities. However, the spatio-temporal graphs among different cities show irregular structures and varied features, which limits the feasibility of existing Few-Shot Learning (FSL) methods. Therefore, we propose a model-agnostic few-shot learning framework for spatio-temporal graph called ST-GFSL. Specifically, to enhance feature extraction by transferring cross-city knowledge, ST-GFSL proposes to generate non-shared parameters based on node-level meta knowledge. The nodes in target city transfer the knowledge via parameter matching, retrieving from similar spatio-temporal characteristics. Furthermore, we propose to reconstruct the graph structure during meta-learning. The graph reconstruction loss is defined to guide structure-aware learning, avoiding structure deviation among different datasets. We conduct comprehensive experiments on four traffic speed prediction benchmarks and the results demonstrate the effectiveness of ST-GFSL compared with state-of-the-art methods.

With the tremendous expansion of graphs data, node classification shows its great importance in many real-world applications. Existing graph neural network based methods mainly focus on classifying unlabeled nodes within fixed classes with abundant labeling. However, in many practical scenarios, graph evolves with emergence of new nodes and edges. Novel classes appear incrementally along with few labeling due to its newly emergence or lack of exploration. In this paper, we focus on this challenging but practical graph few-shot class-incremental learning (GFSCIL) problem and propose a novel method called Geometer. Instead of replacing and retraining the fully connected neural network classifier, Geometer predicts the label of a node by finding the nearest class prototype. Prototype is a vector representing a class in the metric space. With the pop-up of novel classes, Geometer learns and adjusts the attention-based prototypes by observing the geometric proximity, uniformity and separability. Teacher-student knowledge distillation and biased sampling are further introduced to mitigate catastrophic forgetting and unbalanced labeling problem respectively. Experimental results on four public datasets demonstrate that Geometer achieves a substantial improvement of 9.46% to 27.60% over state-of-the-art methods.

Urban traffic flow forecasting is a critical issue in intelligent transportation systems. Due to the complexity and uncertainty of urban road conditions, how to capture the dynamic spatiotemporal correlation and make accurate predictions is very challenging. In most of existing works, urban road network is often modeled as a fixed graph based on local proximity. However, such modeling is not sufficient to describe the dynamics of the road network and capture the global contextual information. In this paper, we consider constructing the road network as a dynamic weighted graph through attention mechanism. Furthermore, we propose to seek both spatial neighbors and semantic neighbors to make more connections between road nodes. We propose a novel Spatiotemporal Adaptive Gated Graph Convolution Network ( STAG-GCN ) to predict traffic conditions for several time steps ahead. STAG-GCN mainly consists of two major components: (1) multivariate self-attention Temporal Convolution Network ( TCN ) is utilized to capture local and long-range temporal dependencies across recent, daily-periodic and weekly-periodic observations; (2) mix-hop AG-GCN extracts selective spatial and semantic dependencies within multi-layer stacking through adaptive graph gating mechanism and mix-hop propagation mechanism. The output of different components are weighted fused to generate the final prediction results. Extensive experiments on two real-world large scale urban traffic dataset have verified the effectiveness, and the multi-step forecasting performance of our proposed models outperforms the state-of-the-art baselines.

Urban traffic flow forecasting is a critical issue in intelligent transportation systems. It is quite challenging due to the complicated spatiotemporal dependency and essential uncertainty brought about by the dynamic urban traffic conditions. In most of existing methods, the spatial correlation is captured by utilizing graph neural networks (GNNs) throughout a fixed graph based on local spatial proximity. However, urban road conditions are complex and changeable, which leads to the interactions between roads should also be dynamic over time. In addition, the global contextual information of roads are also crucial for accurate forecasting. In this paper, we exploit spatiotemporal correlation of urban traffic flow and construct a dynamic weighted graph by seeking both spatial neighbors and semantic neighbors of road nodes. Multi-head self-attention temporal convolution network is utilized to capture local and long-range temporal dependencies across historical observations. Besides, we propose an adaptive graph gating mechanism to extract selective spatial dependencies within multi-layer stacking and correct information deviations caused by artificially defined spatial correlation. Extensive experiments on real world urban traffic dataset from Didi Chuxing GAIA Initiative have verified the effectiveness, and the multi-step forecasting performance of our proposed models outperforms the state-of-the-art baselines. The source code of our model is publicly available at https://github.com/RobinLu1209/STAG-GCN.