Bohan Lin, Kuo Yang, Zelin Tan, Yingchuan Lai, Chen Zhang, Guibin Zhang, Xinlei Yu, Miao Yu, Xu Wang, Yudong Zhang, Yang Wang. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Spatio-temporal (ST) prediction is crucial in earth sciences, including meteorological forecasting and urban computing, to name just a few. Access to ample high-quality data, combined with deep models adept at inference, is essential for attaining significant outcomes. Yet, data scarcity and the substantial costs of sensor deployment result in notable data imbalances. Overly specialized models that lack causal linkages further undermine the generalizability of inference techniques. To address these challenges, we first introduce a causal framework for ST predictions, named $\mathtt{NuwaDynamics}$NuwaDynamicsNuwaDynamics, aimed at pinpointing causal regions in data and providing models with the capability for causal reasoning in a dual-phase process. Initially, we employ upstream self-supervision to identify causally significant patches, equipping the model with generalizable insights and performing targeted interventions on non-essential patches to approximate potential testing distributions. This stage is known as the discovery phase. Progressing from discovery, we apply the insights to downstream tasks tailored to specific ST goals, enhancing the model's recognition of a wider potential data distribution and augmenting its causal perceptual abilities (referred to as the Update phase). Additionally, we address environmental controllability and high computational complexity by implementing channel multiplication and conditional generation methods. This process, termed $\mathtt{NuwaDynamics+}$NuwaDynamics+NuwaDynamics+, can further be interpreted as the front-door adjustment technique in the causality domain. Through comprehensive experiments across ten real-world or simulated ST benchmarks, we demonstrate that integrating the $\mathtt{NuwaDynamics+}$NuwaDynamics+NuwaDynamics+ concept substantially improves various model performance. $\mathtt{NuwaDynamics+}$NuwaDynamics+NuwaDynamics+ concept also significantly enhances the versatility across various dynamic ST tasks, such as extreme weather forecasting and long-temporal-step super-resolution predictions.

Spatiotemporal forecasting (STF) is pivotal in urban computing, yet data scarcity in developing cities hampers robust model training. Addressing this, recent studies leverage transfer learning to migrate knowledge from data-rich (source) to data-poor (target) cities. This strategy, while effective, faces challenges as pre-trained models risk absorbing noise and harmful information due to data distribution disparities, potentially undermining the accuracy of forecasts for target cities. To address this issue, we propose a one-stage STF framework named Target-Skewed Joint Training (TSJT). Central to TSJT is a novel Target-Skewed Backward training strategy that selectively refines gradients from source city data, preserving only the elements that positively impact the target city. To further enhance the quality of these gradients, we have designed a Node Prompting Module (NPM). TSJT is crafted for seamless integration with existing STF models, endowing them with the capability to efficiently tackle challenges stemming from data scarcity. Experimental results on several real-world datasets from multiple cities substantiate the efficacy of TSJT in the realm of cross-city transfer learning.

Typical spatiotemporal graph forecasting methods process graph-structured spatiotemporal data respectively from spatial and temporal perspectives with the idea of divide and conquer. Existing works are incapable of capturing long-term transdimensional correlations among different spatial points in different time planes, i.e., time-space-interlaced correlations. To tackle this issue, we propose a two-stage summarized attention network to establish transdimensional direct message passing routes between different data points in different time planes and spaces, thus enabling the extraction of time-space-interlaced long-term correlations. Specifically, a novel spatiotemporal embedding is proposed to implement time-space-interlaced learning by expanding orthogonal spatial and temporal dimensionalities into one-dimensionality, a series of temporal context fusion units are added to address the fluctuation dislocation insensitivity issue which is caused by time-space-dimension expansion, and an ingenious two-stage design can significantly reduce the computation complexity of such time-space-interlaced learning. Extensive experiments illustrate the superior performance of our proposed approach on real-world spatiotemporal datasets.

Spatiotemporal forecasting facilitates many real world intelligent systems. Combining graph learning with temporal models has recently become popular in spatiotemporal forecasting. Although graph convolution enhances the modeling of spatial correlations, it results in unsatisfactory efficiency and poor extensibility in existing models. Consequently, existing graph-learning-based methods struggle to handle emerging nodes and are difficult to deploy on large-scale datasets. In this paper, we argue that graph learning in spatiotemporal forecasting is to discriminate different nodes thus generating node-specific features. To achieve the same effect of graph learning, we propose to use node embeddings to learn node-specific features and introduce a simple yet well-performing Embedding Enhanced MLP (E2MLP) spatiotemporal forecasting model. A hierarchical MLP framework is proposed, where node embeddings are introduced in each procedure of the framework. E2MLP outperforms existing SOTA spatiotemporal forecasting models on five real-world widely-used datasets, and can be easily scaled to newly emerging nodes without any performance influence on existing nodes. The source codes are publicly available from https://github.com/Ziyan2019/E2MLP

Irregular time series analysis is increasingly essential in data management due to the proliferation of complex data irregularly sampled by real-world systems. Traditional time series models, including RNN-based models and transformer variants, face significant challenges in generalizing to continuous-time paradigms, which are essential for capturing the ongoing dynamics of irregular time series. Neural Ordinary Differential Equations (NODEs) assume a continuous latent dynamic and provide an elegant framework for irregular time series analysis, yet they suffer from limitations like fragmented latent processes and the inability to fully exploit interdependencies among observations. To address these challenges, we propose a novel Differentiable hidden state enhanced neural ODE framework, termed DIFFODE, designed to effectively model irregular time series. Concretely, we introduce an attention-based differential hidden state that maps irregular observations into a continuous hidden state space, enabling the extraction of latent dynamics while preserving temporal continuity. Leveraging the theory of generalized inverses, DIFFODE innovatively derives ODEs to describe hidden state dynamics. Furthermore, we incorporate the Hoyer metric into our framework to enhance its capacity to capture subtle yet critical temporal shifts, significantly improving the accuracy of time series modeling. Extensive experiments on both synthetic and real-world datasets demonstrate the effectiveness of DIFFODE across three key tasks, including irregular time series classification, interpolation, and extrapolation.

Generative flow networks (GFlowNets) have been considered as powerful tools for generating candidates with desired properties. Given that evaluating the property of candidates can be complex and time-consuming, existing GFlowNets train proxy models for efficient online evaluation. However, the performance of proxy models is heavily dependent on the amount of data and is of considerable uncertainty. Therefore, it is of great interest that how to develop an offline GFlowNet that does not rely on online evaluation. Under the offline setting, the limited data results in an insufficient exploration of state space. The insufficient exploration means that offline GFlowNets can hardly generate satisfying candidates out of the distribution of training data. Therefore, it is critical to restrict the offline model to act in the distribution of training data. The distinctive training goal of GFlownets poses a unique challenge for making such restrictions. Tackling the challenge, we propose Conservative Offline GFlowNet (COFlowNet) in this paper. We define unsupported flow, edges containing unseen states in training data. Models can learn extremely little knowledge about unsupported flow from training data. By constraining the model from exploring unsupported flows, we restrict COFlowNet to explore as optimal trajectories on the training set as possible, thus generating better candidates. In order to improve the diversity of candidates, we further introduce a quantile version of unsupported flow restriction. Experimental results on several widely-used datasets validate the effectiveness of COFlowNet in generating high-scored and diverse candidates. All implementations are available at https://github.com/yuxuan9982/COflownet.

Long-term time series forecasting (LTSF) has traditionally relied on large parameters to capture extended temporal dependencies, resulting in substantial computational costs and inefficiencies in both memory usage and processing time. However, time series data, unlike high-dimensional images or text, often exhibit temporal pattern similarity and low-rank structures, especially in long-term horizons. By leveraging this structure, models can be guided to focus on more essential, concise temporal data, improving both accuracy and computational efficiency. In this paper, we introduce TimeBase, an ultra-lightweight network to harness the power of minimalism in LTSF. TimeBase 1) extracts core basis temporal components and 2) transforms traditional point-level forecasting into efficient segment-level forecasting, achieving optimal utilization of both data and parameters. Extensive experiments on diverse real-world datasets show that TimeBase achieves remarkable efficiency and secures competitive forecasting performance. Additionally, TimeBase can also serve as a very effective plug-and-play complexity reducer for any patch-based forecasting models. Code is available at https://github.com/hqh0728/TimeBase.

Discovering regularities from spatiotemporal systems can benefit various scientific and social planning. Current spatiotemporal learners usually train an independent model from a specific source data that leads to limited transferability among sources, where even correlated tasks requires new design and training. The key towards increasing cross-domain knowledge is to enable collective intelligence and model evolution. In this paper, inspired by neuroscience theories, we theoretically derive the increased information boundary via learning cross-domain collective intelligence and propose a Synaptic EVOlutional spatiotemporal network, SynEVO, where SynEVO breaks the model independence and enables cross-domain knowledge to be shared and aggregated. Specifically, we first re-order the sample groups to imitate the human curriculum learning, and devise two complementary learners, elastic common container and task-independent extractor to allow model growth and task-wise commonality and personality disentanglement. Then an adaptive dynamic coupler with a new difference metric determines whether the new sample group should be incorporated into common container to achieve model evolution under various domains. Experiments show that SynEVO improves the generalization capacity by at most 42\% under cross-domain scenarios and SynEVO provides a paradigm of NeuroAI for knowledge transfer and adaptation. Code available at [https://github.com/Rodger-Lau/SynEVO](https://github.com/Rodger-Lau/SynEVO).

Recently, spatiotemporal graph convolutional networks have achieved dominant performance in spatiotemporal prediction tasks. However, most models relying on node-to-node messaging interaction exhibit sensitivity to spatiotemporal shifts, encountering out-of-distribution (OOD) challenges. To address these issues, we introduce \textbf{\underline{S}}patio-\textbf{\underline{T}}emporal \textbf{\underline{O}}OD \textbf{\underline{P}}rocessor (STOP), which employs a centralized messaging mechanism along with a message perturbation mechanism to facilitate robust spatiotemporal interactions. Specifically, the centralized messaging mechanism integrates Context-Aware Units for coarse-grained spatiotemporal feature interactions with nodes, effectively blocking traditional node-to-node messages. We also implement a message perturbation mechanism to disrupt this messaging process, compelling the model to extract generalizable contextual features from generated variant environments. Finally, we customize a spatiotemporal distributionally robust optimization approach that exposes the model to challenging environments, thereby further enhancing its generalization capabilities. Compared with 14 baselines across six datasets, STOP achieves up to \textbf{17.01\%} improvement in generalization performance and \textbf{18.44\%} improvement in inductive learning performance. The code is available at https://github.com/PoorOtterBob/STOP.

Current spatio-temporal modeling techniques largely rely on the abundant data and the design of task-specific models. However, many cities lack well-established digital infrastructures, making data scarcity and the high cost of model development significant barriers to application deployment. Therefore, this work aims to enable spatio-temporal learning to cope with the problems of few-shot data modeling and model generalizability. To this end, we propose a Universal Spatio-Temporal Correlationship pre-training framework (USTC), for spatio-temporal modeling across different cities and tasks. To enhance the spatio-temporal representations during pre-training, we propose to decouple the time-frequency patterns within data, and leverage contrastive learning to maintain the time-frequency consistency. To further improve the adaptability to downstream tasks, we design a prompt generation module to mine personalized spatio-temporal patterns on the target city, which can be integrated with the learned common spatio-temporal representations to collaboratively serve downstream tasks. Extensive experiments conducted on real-world datasets demonstrate that USTC significantly outperforms the advanced baselines in forecasting, imputation, and extrapolation across cities.

The effectiveness of Spatiotemporal Graph Neural Networks (STGNNs) critically hinges on the quality of the underlying graph topology. While end-to-end adaptive graph learning methods have demonstrated promising results in capturing latent spatiotemporal dependencies, they often suffer from high computational complexity and limited expressive capacity. In this paper, we propose MAGE for efficient spatiotemporal forecasting. We first conduct a theoretical analysis demonstrating that the ReLU activation function employed in existing methods amplifies edge-level noise during graph topology learning, thereby compromising the fidelity of the learned graph structures. To enhance model expressiveness, we introduce a sparse yet balanced mixture-of-experts strategy, where each expert perceives the unique underlying graph through kernel-based functions and operates with linear complexity relative to the number of nodes. The sparsity mechanism ensures that each node interacts exclusively with compatible experts, while the balancing mechanism promotes uniform activation across all experts, enabling diverse and adaptive graph representations. Furthermore, we theoretically establish that a single graph convolution using the learned graph in MAGE is mathematically equivalent to multiple convolutional steps under conventional graphs. We evaluate MAGE against advanced baselines on multiple real-world spatiotemporal datasets. MAGE achieves competitive performance while maintaining strong computational efficiency.

While existing spatiotemporal prediction models have shown promising performance, they often rely on the assumption of input-label spatiotemporal consistency, and their high complexity raises concerns about scalability. To enhance both efficiency and performance, we integrate label information into the learning process and propose a spatiotemporal dynamic theory that outlines a bi-directional learning paradigm. Building on this paradigm, we design BiST, a lightweight yet effective Bi -directional S patio -T emporal prediction model. BiST incorporates two key processes: a forward spatiotemporal learning process and a backward correction process. The forward process utilizes MLP layers exclusively to model input correlations and generate base prediction. In the backward process, we implement a spatiotemporal decoupling module, which can learn the residual modeling deviation between input and label representations from a decoupled perspective. After smoothing the residual with a diffusion module, we can obtain the correction term to correct the base predictions. This innovative design enables BiST to achieve competitive performance while remaining lightweight. We evaluate BiST against 26 baselines across 13 datasets, including a large-scale dataset with ten thousand nodes and a longrange dataset spanning 20 years. An impressive experimental result demonstrates that BiST achieves a 8.13% improvement in performance compared to state-of-the-art models while consuming only 1.86% of the training time and 7.36% of the memory usage.

Subgraph learning has dominated most practices of improving the expressive power of Message Passing Neural Networks (MPNNs). Existing subgraph discovery policies can be classified into node-based and partition-based, which both achieve impressive performance in most scenarios. However, both mainstream solutions still face a subgraph degradation trap. Subgraph degradation is reflected in the phenomenon that the subgraph-level methods fail to offer any benefits over node-level MPNNs. In this work, we empirically investigate the existence of the subgraph degradation issue and introduce a unified perspective, perfect reconstruction, to provide insights for improving two lines of methods. We further propose a subgraph learning strategy guided by the principle of perfect reconstruction. To achieve this, two major issues should be well-addressed, i.e.,(i) how to ensure the subgraphs to possess with ‘perfect’ information? (ii) how to guarantee the ‘reconstruction’ power of obtained subgraphs?First, we propose a subgraph partition strategyRayleigh-resistanceto extract non-overlap subgraphs by leveraging the graph spectral theory. Second, we put forward aQuerymechanism to achieve subgraph-level equivariant learning, which guarantees subgraph reconstruction ability. These two parts,perfect subgraph partitionandequivariant subgraph learningare seamlessly unified as a novelRayleigh-resistanceEquivariantSubgraph learningarchitecture (RayE-Sub). Comprehensive experiments on both synthetic and real datasets demonstrate that our approach can consistently outperform previous subgraph learning architectures.

Pre-training GNNs to extract transferable knowledge and apply it to downstream tasks has become the de facto standard of graph representation learning. Recent works focused on designing self-supervised pre-training tasks to extract useful and universal transferable knowledge from large-scale unlabeled data. However, they have to face an inevitable question: traditional pre-training strategies that aim at extracting useful information about pre-training tasks, may not extract all useful information about the downstream task. In this paper, we reexamine the pre-training process within traditional pre-training and fine-tuning frameworks from the perspective of Information Bottleneck (IB) and confirm that the forgetting phenomenon in pre-training phase may cause detrimental effects on downstream tasks. Therefore, we propose a novelDelayedBottleneckingPre-training (DBP) framework which maintains as much as possible mutual information between latent representations and training data during pre-training phase by suppressing the compression operation and delays the compression operation to fine-tuning phase to make sure the compression can be guided with labeled fine-tuning data and downstream tasks. To achieve this, we design two information control objectives that can be directly optimized and further integrate them into the actual model design. Extensive experiments on both chemistry and biology domains demonstrate the effectiveness of DBP.

Understanding mobile traffic data and predicting future trends are essential for wireless operators and service providers to allocate resources efficiently and manage energy effectively. Despite the strong performance of existing models, accurately forecasting mobile traffic remains a challenge due to limited spatial and temporal modeling capabilities and high computational complexity. This paper introduces MobiMixer, a lightweight and efficient multi-scale spatiotemporal mixing model. Its core concept is to integrate multi-scale information from both spatial and temporal dimensions to improve performance on mobile traffic data. We develop a hierarchical interaction module that incorporates super nodes to enable global high-level feature interactions among nodes with common patterns. Additionally, we employ a dynamic time warping strategy to decouple mobile traffic sequences into stable and seasonal components, which are then modeled at different scales using a multi-scale temporal mixing module. We conduct extensive experiments on mobile traffic datasets collected from four international cities. Compared with 21 state-of-the-art benchmark models, MobiMixer demonstrates highly competitive performance, achieving a maximum improvement of 48.49% on the Milan mobile dataset. The model achieves an improvement in training efficiency of up to 10.69 times and reduces memory usage by 33.01%. The source code is available athttps://github.com/PoorOtterBob/Submitted_Code.

With the progress of urban transportation systems, a significant amount of high-quality traffic data is continuously collected through streaming manners, which has propelled the prosperity of the field of spatial-temporal graph prediction. In this paper, rather than solely focusing on designing powerful models for static graphs, we shift our focus to spatial-temporal graph prediction in the dynamic scenario, which involves a continuously expanding and evolving underlying graph. To address inherent challenges, a decoupled learning framework (DLF) is proposed in this paper, which consists of a spatial-temporal graph learning network (DSTG) with a specialized decoupling training strategy. Incorporating inductive biases of time-series structures, DSTG can interpret time dependencies into latent trend and seasonal terms. To enable prompt adaptation to the evolving distribution of the dynamic graph, our decoupling training strategy is devised to iteratively update these two types of patterns. Specifically, for learning seasonal patterns, we conduct thorough training for the model using a long time series (e.g., three months of data). To enhance the learning ability of the model, we also introduce the masked auto-encoding mechanism. During this period, we frequently update trend patterns to expand new information from dynamic graphs. Considering both effectiveness and efficiency, we develop a subnet sampling strategy to select a few representative nodes for fine-tuning the weights of the model. These sampled nodes cover unseen patterns and previously learned patterns. Experiments on dynamic spatial-temporal graph datasets further demonstrate the competitive performance, superior efficiency, and strong scalability of the proposed framework.

One of the most important tasks in ride-hailing is order dispatching, i.e., assigning unserved orders to available drivers. Recent order dispatching has achieved a significant improvement due to the advance of reinforcement learning, which has been approved to be able to effectively address sequential decision-making problems like order dispatching. However, most existing reinforcement learning methods require agents to learn the optimal policy by interacting with environments online, which is challenging or impractical for real-world deployment due to high costs or safety concerns. For example, due to the spatiotemporally unbalanced supply and demand, online reinforcement learning-based order dispatching may significantly impact the revenue of the ride-hailing platform and passenger experience during the policy learning period. Hence, in this work, we develop an offline deep reinforcement learning framework called NondBREM for large-scale order dispatching, which learns policy from only the accumulated logged data to avoid costly and unsafe interactions with the environment. In NondBREM, a Nondeterministic Batch-Constrained Q-learning (NondBCQ) module is developed to reduce the algorithm extrapolation error and a Random Ensemble Mixture (REM) module that integrates multiple value networks with multi-head networks is utilized to improve the model generalization and robustness. Extensive experiments on large-scale real-world ride-hailing datasets show the superiority of our design.

Seeking for generalizable graph representations becomes hot spot in the area of graph learning. Recently, causality theory has been applied for extracting the causal relations between graph data and labels, which are generalizable under distribution shift and result in better OOD generalization. In this paper, for more accurately capturing causal representation of graph data, we propose a gradient reactivation enhanced causal subgraph extraction method. The proposed model utilizes attention mechanism to extract the causal features and attenuates the confounding effect of shortcut features. For ensuring stability of extracted causal features, we propose a novel gradient reactivation method to filter features with greater effect on making prediction. Extensively experimental result proves the effectiveness of the proposed model.

Real-world tasks are universally associated with training samples that exhibit a long-tailed class distribution, and traditional deep learning models are not suitable for fitting this distribution, thus resulting in a biased trained model. To surmount this dilemma, massive deep long-tailed learning studies have been proposed to achieve inter-class fairness models by designing sophisticated sampling strategies or improving existing model structures and loss functions. Habitually, these studies tend to apply data augmentation strategies to improve the generalization performance of their models. However, this augmentation strategy applied to balanced distributions may not be the best option for long-tailed distributions. For a profound understanding of data augmentation, we first theoretically analyze the gains of traditional augmentation strategies in long-tailed learning, and observe that augmentation methods cause the long-tailed distribution to be imbalanced again, resulting in an intertwined imbalance: inherent data-wise imbalance and extrinsic augmentation-wise imbalance, i.e., two 'birds' co-exist in long-tailed learning. Motivated by this observation, we propose an adaptive Dynamic Optional Data Augmentation (DODA) to address this intertwined imbalance, i.e., one 'stone' simultaneously 'kills' two 'birds', which allows each class to choose appropriate augmentation methods by maintaining a corresponding augmentation probability distribution for each class during training. Extensive experiments across mainstream long-tailed recognition benchmarks (e.g., CIFAR-100-LT, ImageNet-LT, and iNaturalist 2018) prove the effectiveness and flexibility of the DODA in overcoming the intertwined imbalance.

Traffic prediction is a crucial task in the Intelligent Transportation System (ITS), receiving significant attention from both industry and academia. Numerous spatio-temporal graph convolutional networks have emerged for traffic prediction and achieved remarkable success. However, these models have limitations in terms of generalization and scalability when dealing with Out-of-Distribution (OOD) graph data with both structural and temporal shifts. To tackle the challenges of spatio-temporal shift, we propose a framework called STONE by learning invariable node dependencies, which achieve stable performance in variable environments. STONE initially employs gated-transformers to extract spatial and temporal semantic graphs. These two kinds of graphs represent spatial and temporal dependencies, respectively. Then we design three techniques to address spatio-temporal shifts. Firstly, we introduce a Fréchet embedding method that is insensitive to structural shifts, and this embedding space can integrate loose position dependencies of nodes within the graph. Secondly, we propose a graph intervention mechanism to generate multiple variant environments by perturbing two kinds of semantic graphs without any data augmentations, and STONE can explore invariant node representation from environments. Finally, we further introduce an explore-to-extrapolate risk objective to enhance the variety of generated environments. We conduct experiments on multiple traffic datasets, and the results demonstrate that our proposed model exhibits competitive performance in terms of generalization and scalability.

Spatiotemporal traffic forecasting plays a critical role in intelligent transportation systems, which empowers diverse urban services. Existing traffic forecasting frameworks usually devise various learning strategies to capture spatiotemporal correlations from the perspective of volume itself. However, we argue that previous traffic predictions are still unreliable due to two aspects. First, the influences of context factor-wise interactions on dynamic region-wise correlations are under exploitation. Second, the dynamics induce the credibility issue of forecasting that has not been well-explored. In this paper, we exploit the informative traffic-related context factors to jointly tackle the dynamic regional heterogeneity and explain the stochasticity, towards a credible uncertainty-aware traffic forecasting. Specifically, to internalize the dynamic contextual influences into learning process, we design a context-cross relational embedding to capture interactions between each context, and generate virtual graph topology to dynamically relate pairwise regions with context embedding. To quantify the prediction credibility, we attribute data-side aleatoric uncertainty to contexts and re-utilize them for aleatoric uncertainty quantification. Then we couple a dual-pipeline learning with the same objective to produce the discrepancy of model outputs and quantify model-side epistemic uncertainty. These two uncertainties are fed through a spatiotemporal network for extracting uncertainty evolution patterns. Finally, comprehensive experiments and model deployments have corroborated the credibility of our framework.

Graph Neural Networks (GNNs) can learn representative graph-level features to achieve efficient graph classification. But GNNs usually assume an environment where both class and structure distribution are balanced. Although previous works have considered the graph classification problem under the scenario of class imbalance or structure imbalance, they habitually ignored the obvious fact that class imbalance and structural imbalance are often intertwined in the real world. In this paper, we propose a carefully designed structure-driven learning framework called ImbGNN to address the potential intertwined class imbalance and structural imbalance in graph classification. Specifically, we find that feature-oriented augmentation (e.g., feature masking) and structure-oriented augmentation (e.g., edge perturbation) will have differential impacts when applied to different graphs. Therefore, we design optional augmentation based on the average degree distribution to alleviate structural imbalance. Furthermore, based on the imbalance of graph size distribution, we utilize a similarity-friendly graph random walk to extract a core subgraph to improve the accuracy of graph kernel similarity calculation, and then construct a more reasonable kernel-based graph of graphs, thereby alleviating the class imbalance and size imbalance. Extensive experiments on multiple benchmark datasets demonstrate that our proposed ImbGNN framework outperforms previous baselines on imbalanced graph classification tasks. The code of ImbGNN is available in~https://github.com/Xiaovy/ImbGNN.

Traffic forecasting is a challenging research topic due to the complex spatial and temporal dependencies among different roads. Though great efforts have been made on traffic forecasting, existing works still have the following shortcomings: i) Most methods only directly perform on the original road network topology which cannot accommodate the diverse traffic patterns and multi-granularity traffic forecasting requirements driven by the natural multi-level urban structure and layout, ii) The existing studies based on the spatio-temporal multi-granularity perspective ignore the interactions between the fine-grained information and coarse-grained information, resulting in the spatio-temporal correlation under multi-granularity inaccurately modeled. To solve the problems, we propose an Adaptive and Interactive Multi-level Spatio-Temporal network (AIMST) for traffic forecasting. Specifically, we first devise a learnable adaptive hierarchical clustering method to automatically generate more coarse-grained graphs from the initial road networks and the traffic data. Then, the spatio-temporal graph convolutional networks are executed on the constructed hierarchical traffic graph of each level correspondingly to capture the spatio-temporal patterns. Furthermore, a multi-level bidirectional interaction module is designed to emphasize the multi-grained interaction patterns among different levels. Extensive experiments on two real-world traffic datasets demonstrate that our framework is superior to several state-of-the-art baselines.

Spatio-temporal mobility modeling plays a pivotal role in the advancement of mobile computing. Nowadays, data is frequently held by various distributed silos, which are isolated from each other and confront limitations on data sharing. Given this, there have been some attempts to introduce federated learning into spatio-temporal mobility modeling. Meanwhile, the distributional heterogeneity inherent in the spatio-temporal data also puts forward requirements for model personalization. However, the existing methods tackle personalization in a model-centric manner and fail to explore the data characteristics in various data silos, thus ignoring the fact that the fundamental cause of insufficient personalization in the model is the heterogeneous distribution of data. In this paper, we propose a novel distribution-oriented personalizedFederated learning framework forCross-siloSpatio-Temporal mobility modeling (namedFedCroST), that leverages learnable spatio-temporal prompts to implicitly represent the local data distribution patterns of data silos and guide the local models to learn the personalized information. Specifically, we focus on the potential characteristics within temporal distribution and devise a conditional diffusion module to generate temporal prompts that serve as guidance for the evolution of the time series. Simultaneously, we emphasize the structure distribution inherent in node neighborhoods and propose adaptive spatial structure partition to construct the spatial prompts, augmenting the spatial information representation. Furthermore, we introduce a denoising autoencoder to effectively harness the learned multi-view spatio-temporal features and obtain personalized representations adapted to local tasks. Our proposal highlights the significance of latent spatio-temporal data distributions in enabling personalized federated spatio-temporal learning, providing new insights into modeling spatio-temporal mobility in data silo scenarios. Extensive experiments conducted on real-world datasets demonstrate that FedCroST outperforms the advanced baselines by a large margin in diverse cross-silo spatio-temporal mobility modeling tasks.

Object detection is a fundamental task in computer vision, consisting of both classification and localization tasks. Previous works mostly perform classification and localization with shared feature extractor like Convolution Neural Network. However, the tasks of classification and localization exhibit different sensitivities with regard to the same feature, hence the "task spatial misalignment" issue. This issue can result in a hedge issue between the performances of localizer and classifier. To address these issues, we first propose a novel Dynamic Coefficient Loss to simultaneously consider and balance the performances of classification and localization tasks. To well address anchor label misjudgment issue in irregular- shaped object detection, we define a new classification-aware IoU metric to assign anchors intelligently. Finally, we further introduce the localization factor into NMS by proposing a Classification-Localization balanced NMS. Extensive experiments on MS COCO and PASCAL VOC demonstrate that our proposals can improve RetinaNet by around 1.5% AP with various backbones.

Recently, spatiotemporal graph convolutional networks are becoming popular in the field of traffic flow prediction and significantly improve prediction accuracy. However, the majority of existing traffic flow prediction models are tailored to static traffic networks and fail to model the continuous evolution and expansion of traffic networks. In this work, we move to investigate the challenge of traffic flow prediction on an expanding traffic network. And we propose an efficient and effective continual learning framework to achieve continuous traffic flow prediction without the access to historical graph data, namely Pattern Expansion and Consolidation based on Pattern Matching based (PECPM). Specifically, we first design a pattern bank based on pattern matching to store representative patterns of the road network. With the expansion of the road network, the model configured with such a bank module can achieve continuous traffic prediction by effectively managing patterns stored in the bank. The core idea is to continuously update new patterns while consolidating learned ones. Specifically, we design a pattern expansion mechanism that can detect evolved and new patterns from the updated network, then these unknown patterns are expanded into the pattern bank to adapt to the updated road network. Additionally, we propose a pattern consolidation mechanism that includes both a bank preservation mechanism and a pattern traceability mechanism. This can effectively consolidate the learned patterns in the bank without requiring access to detailed historical graph data. We construct experiments on real-world traffic datasets to demonstrate the competitive performance, superior efficiency, and strong generalization ability of PECPM.

Missing values, which are common in multivariate time series, is most important obstacle towards the utilization and interpretation of those data. Great efforts have been employed on how to accurately impute missing values in multivariate time series, and existing works either use deep learning networks to achieve deterministic imputations or aim at generating different plausible imputations by sampling multiple noises from a same distribution and then denoising them. However, these models either fall short of modeling the uncertainties of imputations due to their deterministic nature or perform poorly in terms of interpretability and imputation accuracy due to their ignorance of the correlations between the latent representations of both observed and missing values which are parts of samples from a same distribution. To this end, in this paper, we explicitly take the correlations between observed and missing values into account, and theoretically re-derive the Evidence Lower BOund (ELBO) of conditional diffusion model in the scenario of multivariate time series imputation. Based on the newly derived ELBO, we further propose a novel multivariate imputation diffusion model (MIDM) which is equipped with novel noise sampling, adding and denoising mechanisms for multivariate time series imputation, and the series of newly designed technologies jointly ensure the involving of the consistency between observed and missing values. Extensive experiments on both the tasks of multivariate time series imputation and forecasting witness the superiority of our proposed MIDM model on generating conditional estimations.

Subgraph learning has received considerable attention in its capacity of interpreting important structural information for predictions. Existing subgraph learning usually exploits statistics on predefined structures e.g., node degrees, occurrence frequency, to extract subgraphs, or refine the contents via only capturing label-relevant information with node-level sampling. Given diverse subgraph patterns, and mutual independence with local correlations on graphs, current solutions on subgraph learning still have two limitations in extraction and refinement stages. 1) The universality of extracting substructure patterns across domains is still lacking, 2) node-level sampling in refinement will distort the original local topology and none explicit guidance eliminating redundant information contribute to inefficiency issue. In this paper, we propose a unified subgraph learning scheme, Poly-Pivot Graph Neural Network (P2GNN) where we designate the centric node of each subgraph as the pivot. In the extraction stage, we present a general subgraph extraction principle, i.e., Local; Asymmetry between the centric and affiliated nodes. To this end, we asymmetrically model the similarity between each pair of nodes with random walk and quantify mutual affiliations in Affinity Propagation architecture, to extract subgraph structures. In the refinement, we devise a subgraph-level exclusion regularization to squash the target-independent information by considering mutual relations across subgraphs, cooperatively preserving a support set of subgraphs and facilitating the refinement process for graph representation. Empirical experiments on diverse web and biological graphs reveal 1.1%~7.3% improvements against best baselines, and visualized case studies prove the universality and interpretability of our P2GNN.

Spatiotemporal (ST) learning has become a crucial technique for urban digitalization. Due to expansions and dynamics of cities, current spatiotemporal models are inclined to suffer distribution shifts between training and testing sets, leading to the OOD delimma. However, few studies focus on such OOD problem in temporal regressions, let alone spatiotemporal learning. Spatiotemporal data usually reveals segment-level heterogeneity within periodicity and complex spatial dependencies, posing challenges to invariance extraction. In this paper, we find that ST relations make sense for generalization and devise a Causal ST learning framework, CauSTG, which enables invariant relation transferred to OOD scenarios. Specifically, we take temporal steps as environments, and transform spatial-temporal relations into learnable parameters. To tackle heterogeneity in periodicity, we partition temporal steps into sub-environments by identifying distinctive trend patterns, enabling re-organized samples trained separately. To extract invariance within ST observations, we propose a spatiotemporal consistency learner and a hierarchical invariance explorer to jointly filter out stable relations. Our spatiotemporal learner quantifies bi-directional spatial consistency and extracts disentangled seasonal-trend patterns via trainable parameters. Further, the hierarchical invariance explorer constructs variation-based filter to achieve both local and global invariances. Experiments reveal that CauSTG can increase at most 10.26% performance against best baselines, and visualized invariant relations can well interpret the physical rationales. The appendix and codes can be available in our Github repository.

Spatiotemporal data forecasting is a fundamental task in the field of graph data mining. Typical spatiotemporal data prediction methods usually capture spatial dependencies by directly aggregating features of local neighboring vertices in a fixed graph. However, this kind of aggregators can only capture localized correlations between vertices, and while been stacked for larger receptive field, they fall into the dilemma of over-smoothing. Additional, in temporal perspective, traditional methods focus on fixed graphs, while the correlations among vertexes can be dynamic. And time series components integrated strategies in traditional spatiotemporal learning methods can hardly handle frequently and drastically changed sequences. To overcome those limitations of existing works, in this paper, we propose a novel multi-graph based dynamic learning framework. First, a novel Dynamic Neighbor Search (DNS) mechanism is introduced to model global dynamic correlations between vertices by constructing a feature graph (FG), where the adjacency matrix is dynamically determined by DNS. Then we further alleviate the over-smoothing issue with our newly designed Adaptive Heterogeneous Representation (AHR) module. Both FG and origin graph (OG) are fed into the AHR modules and fused in our proposed Multi-graph Fusion block. Additionally, we design a Differential Vertex Representation (DVR) module which takes advantage of differential information to model temporal trends. Extensive experiments illustrate the superior forecasting performances of our proposed multi-graph based dynamic learning framework on six real-world spatiotemporal datasets from different cities and domains, and this corroborates the solid effectiveness of our proposed framework and its superior generalization ability.

Accurate traffic prediction plays a vital role in intelligent transport managements and applications. However, in the vast majority of existing works, the focus is mainly on modeling spatiotemporal correlations in static traffic networks. Thus, the continuous expansion and evolution of traffic networks are ignored. In this work, we study the problem of traffic prediction with expanding road network structures under the continual learning paradigm. Considering the model prediction performance, efficiency, and data accessibility, a SpatioTemporal Knowledge Expansion and Consolidation (STKEC) framework is proposed. This framework contains an influence-based knowledge expansion strategy to help the spatiotemporal learning model integrate new spatiotemporal traffic patterns and a memory-augmented knowledge consolidation mechanism to preserve the learned spatiotemporal patterns without accessing the data in previous graphs. Extensive experiments are conducted on a large-scale dataset and verify the superior performance of STKEC in continual traffic prediction.

Urban housing price is widely accepted as an economic indicator which is of both business and research interest in urban computing. However, due to the complex nature of influencing factors and the sparse property of transaction records, to implement such a model is still challenging. To address these challenges, in this work, we study an effective and fine-grained model for urban subregion housing price predictions. Compared to existing works, our proposal improves the forecasting granularity from city-level to mile-level, with only publicly released transaction data. We employ a feature selection mechanism to select more relevant features. Then, we propose an integrated model, JGC MMN (Joint Gated Co-attention Based Multi-modal Network), to learn all-level features and capture spatiotemporal correlations in all-time stages with a modified densely connected convolutional network as well as current ingredients and future expectations. Next, we devise a novel JGC based fusion method to better fuse the heterogeneous data of multi-stage models by considering their interactions in temporal dimension. Finally, extensive empirical studies on real datasets demonstrate the effectiveness of our proposal, and this fine-grained housing price forecasting has the potential to support a broad scope of applications, ranging from urban planning to housing market recommendations.

Modern intelligent transportation system (ITS) has greatly benefitted people's daily life. However, the chanciness and suddenness of urban anomalies may greatly restrict the trouble-free operations of ITS. To be aware of future urban anomalies and their possible influences, great efforts have been achieved on these two aspects, but comprehensive predictions of urban anomalies including the predictions of distributions and durations, are still beingless. And the spatiotemporal cascade self/mutual exciting influences among anomalies have never been considered in previous studies. In this paper, we propose a novel Anomaly Distribution and Duration Joint-Prediction (A2DJP) algorithm to simultaneously filtrate urban subregions and estimate the duration of corresponding potential anomalies in the future. To capture the spatiotemporal correlations between urban traffics and anomalies, we use a modified Graph Convolution Network and Long Short-Term Memory integrated network. To learn the cascade correlations among anomalies themselves, we devise a novel Spatiotemporal neural Hawkes Process model, which contains a Hawkes Process (HP) based GCN and HP-based LSTM to extract the anomaly-wise spatiotemporal cascading correlations. By fusing the spatiotemporal correlations between traffics and anomalies, we then simultaneously predict the distributions and durations of future anomalies. Extensive experiments on real-world datasets demonstrate that our proposed method significantly outperforms state-of-the-art solutions.

Fusing multimodal heterogeneous data plays a vital role in recognition and prediction tasks in various fields, e.g., action recognition and traffic accident forecast. Yet, there remain some key challenges, such as heterogeneous feature interaction and feature redundancies, that significantly affect the performance of multimodal fusion. To tackle these challenges, we first devise a Unified Feature Interaction Module (UFIM) in which a novel orthogonal attention component is designed to obtain fine-grained inter-modal interaction information among heterogeneous features. Then, we propose a novel Self-Correcting Transformer Module (SCTM) which employs a modified transformer to obtain the one-to-many correlation information between the current modal feature and the merged features of other modalities to alleviate the redundancy problem. Extensive experiments on four cross-domain tasks demonstrate the effectiveness and generalization ability of our proposed method.

Since electroencephalogram (EEG) signals can truly reflect human emotional state, emotion recognition based on EEG has turned into a critical branch in the field of artificial intelligence. Aiming at the disparity of EEG signals in various emotional states, we propose a new deep learning model named three-dimension convolution attention neural network (3DCANN) for EEG emotion recognition in this paper. The 3DCANN model is composed of spatio-temporal feature extraction module and EEG channel attention weight learning module, which can extract the dynamic relation well among multi-channel EEG signals and the internal spatial relation of multi-channel EEG signals during continuous period time. In this model, the spatio-temporal features are fused with the weights of dual attention learning, and the fused features are input into the softmax classifier for emotion classification. In addition, we utilize SJTU Emotion EEG Dataset (SEED) to appraise the feasibility and effectiveness of the proposed algorithm. Finally, experimental results display that the 3DCANN method has superior performance over the state-of-the-art models in EEG emotion recognition.

Affective computing is one of the key technologies to achieve advanced brain-machine interfacing. It is increasingly concerning research orientation in the field of artificial intelligence. Emotion recognition is closely related to affective computing. Although emotion recognition based on electroencephalogram (EEG) has attracted more and more attention at home and abroad, subject-independent emotion recognition still faces enormous challenges. We proposed a subject-independent emotion recognition algorithm based on dynamic empirical convolutional neural network (DECNN) in view of the challenges. Combining the advantages of empirical mode decomposition (EMD) and differential entropy (DE), we proposed a dynamic differential entropy (DDE) algorithm to extract the features of EEG signals. After that, the extracted DDE features were classified by convolutional neural networks (CNN). Finally, the proposed algorithm is verified on SJTU Emotion EEG Dataset (SEED). In addition, we discuss the brain area closely related to emotion and design the best profile of electrode placements to reduce the calculation and complexity. Experimental results show that the accuracy of this algorithm is 3.53 percent higher than that of the state-of-the-art emotion recognition methods. What's more, we studied the key electrodes for EEG emotion recognition, which is of guiding significance for the development of wearable EEG devices.