Hierarchical Multi-modal Transformer for Cross-modal Long Document Classification

Long Document Classification (LDC) has gained significant attention recently. However, multi-modal data in long documents such as texts and images are not being effectively utilized. Prior studies in this area have attempted to integrate texts and images in document-related tasks, but they have only focused on short text sequences and images of pages. How to classify long documents with hierarchical structure texts and embedding images is a new problem and faces multi-modal representation difficulties. In this paper, we propose a novel approach called Hierarchical Multi-modal Transformer (HMT) for cross-modal long document classification. The HMT conducts multi-modal feature interaction and fusion between images and texts in a hierarchical manner. Our approach uses a multi-modal transformer and a dynamic multi-scale multi-modal transformer to model the complex relationships between image features, and the section and sentence features. Furthermore, we introduce a new interaction strategy called the dynamic mask transfer module to integrate these two transformers by propagating features between them. To validate our approach, we conduct cross-modal LDC experiments on two newly created and two publicly available multi-modal long document datasets, and the results show that the proposed HMT outperforms state-of-the-art single-modality and multi-modality methods.

Unleash Graph Neural Networks from Heavy Tuning

Graph Neural Networks (GNNs) are deep-learning architectures designed for graph-type data, where understanding relationships among individual observations is crucial. However, achieving promising GNN performance, especially on unseen data, requires comprehensive hyperparameter tuning and meticulous training. Unfortunately, these processes come with high computational costs and significant human effort. Additionally, conventional searching algorithms such as grid search may result in overfitting on validation data, diminishing generalization accuracy. To tackle these challenges, we propose a graph conditional latent diffusion framework (GNN-Diff) to generate high-performing GNNs directly by learning from checkpoints saved during a light-tuning coarse search. Our method: (1) unleashes GNN training from heavy tuning and complex search space design; (2) produces GNN parameters that outperform those obtained through comprehensive grid search; and (3) establishes higher-quality generation for GNNs compared to diffusion frameworks designed for general neural networks.

ST-MambaSync: The Complement of Mamba and Transformers for Spatial-Temporal in Traffic Flow Prediction

Accurate traffic flow prediction is crucial for optimizing traffic management, enhancing road safety, and reducing environmental impacts. Existing models face challenges with long sequence data, requiring substantial memory and computational resources, and often suffer from slow inference times due to the lack of a unified summary state. This paper introduces ST-MambaSync, an innovative traffic flow prediction model that combines transformer technology with the ST-Mamba block, representing a significant advancement in the field. We are the pioneers in employing the Mamba mechanism which is an attention mechanism integrated with ResNet within a transformer framework, which significantly enhances the model's explainability and performance. ST-MambaSync effectively addresses key challenges such as data length and computational efficiency, setting new benchmarks for accuracy and processing speed through comprehensive comparative analysis. This development has significant implications for urban planning and real-time traffic management, establishing a new standard in traffic flow prediction technology.

ST-MambaSync: The Confluence of Mamba Structure and Spatio-Temporal Transformers for Precipitous Traffic Prediction

Balancing accuracy with computational efficiency is paramount in machine learning, particularly when dealing with high-dimensional data, such as spatial-temporal datasets. This study introduces ST-MambaSync, an innovative framework that integrates a streamlined attention layer with a simplified state-space layer. The model achieves competitive accuracy in spatial-temporal prediction tasks. We delve into the relationship between attention mechanisms and the Mamba component, revealing that Mamba functions akin to attention within a residual network structure. This comparative analysis underpins the efficiency of state-space models, elucidating their capability to deliver superior performance at reduced computational costs.

ST-SSMs: Spatial-Temporal Selective State of Space Model for Traffic Forecasting

Accurate and efficient traffic prediction is crucial for planning, management, and control of intelligent transportation systems. Most state-of-the-art methods for traffic prediction effectively predict both long-term and short-term by employing spatio-temporal neural networks as prediction models, together with transformers to learn global information on prediction objects (e.g., traffic states of road segments). However, these methods often have a high computational cost to obtain good performance. This paper introduces an innovative approach to traffic flow prediction, the Spatial-Temporal Selective State Space Model (ST-SSMs), featuring the novel ST-Mamba block, which can achieve good prediction accuracy with less computational cost. A comparative analysis highlights the ST-Mamba layer's efficiency, revealing its equivalence to three attention layers, yet with markedly reduced processing time. Through rigorous testing on diverse real-world datasets, the ST-SSMs model demonstrates exceptional improvements in prediction accuracy and computational simplicity, setting new benchmarks in the domain of traffic flow forecasting

CCDSReFormer: Traffic Flow Prediction with a Criss-Crossed Dual-Stream Enhanced Rectified Transformer Model

Accurate, and effective traffic forecasting is vital for smart traffic systems, crucial in urban traffic planning and management. Current Spatio-Temporal Transformer models, despite their prediction capabilities, struggle with balancing computational efficiency and accuracy, favoring global over local information, and handling spatial and temporal data separately, limiting insight into complex interactions. We introduce the Criss-Crossed Dual-Stream Enhanced Rectified Transformer model (CCDSReFormer), which includes three innovative modules: Enhanced Rectified Spatial Self-attention (ReSSA), Enhanced Rectified Delay Aware Self-attention (ReDASA), and Enhanced Rectified Temporal Self-attention (ReTSA). These modules aim to lower computational needs via sparse attention, focus on local information for better traffic dynamics understanding, and merge spatial and temporal insights through a unique learning method. Extensive tests on six real-world datasets highlight CCDSReFormer's superior performance. An ablation study also confirms the significant impact of each component on the model's predictive accuracy, showcasing our model's ability to forecast traffic flow effectively.

IME: Integrating Multi-curvature Shared and Specific Embedding for Temporal Knowledge Graph Completion

Temporal Knowledge Graphs (TKGs) incorporate a temporal dimension, allowing for a precise capture of the evolution of knowledge and reflecting the dynamic nature of the real world. Typically, TKGs contain complex geometric structures, with various geometric structures interwoven. However, existing Temporal Knowledge Graph Completion (TKGC) methods either model TKGs in a single space or neglect the heterogeneity of different curvature spaces, thus constraining their capacity to capture these intricate geometric structures. In this paper, we propose a novel Integrating Multi-curvature shared and specific Embedding (IME) model for TKGC tasks. Concretely, IME models TKGs into multi-curvature spaces, including hyperspherical, hyperbolic, and Euclidean spaces. Subsequently, IME incorporates two key properties, namely space-shared property and space-specific property. The space-shared property facilitates the learning of commonalities across different curvature spaces and alleviates the spatial gap caused by the heterogeneous nature of multi-curvature spaces, while the space-specific property captures characteristic features. Meanwhile, IME proposes an Adjustable Multi-curvature Pooling (AMP) approach to effectively retain important information. Furthermore, IME innovatively designs similarity, difference, and structure loss functions to attain the stated objective. Experimental results clearly demonstrate the superior performance of IME over existing state-of-the-art TKGC models.

Design Your Own Universe: A Physics-Informed Agnostic Method for Enhancing Graph Neural Networks

Physics-informed Graph Neural Networks have achieved remarkable performance in learning through graph-structured data by mitigating common GNN challenges such as over-smoothing, over-squashing, and heterophily adaption. Despite these advancements, the development of a simple yet effective paradigm that appropriately integrates previous methods for handling all these challenges is still underway. In this paper, we draw an analogy between the propagation of GNNs and particle systems in physics, proposing a model-agnostic enhancement framework. This framework enriches the graph structure by introducing additional nodes and rewiring connections with both positive and negative weights, guided by node labeling information. We theoretically verify that GNNs enhanced through our approach can effectively circumvent the over-smoothing issue and exhibit robustness against over-squashing. Moreover, we conduct a spectral analysis on the rewired graph to demonstrate that the corresponding GNNs can fit both homophilic and heterophilic graphs. Empirical validations on benchmarks for homophilic, heterophilic graphs, and long-term graph datasets show that GNNs enhanced by our method significantly outperform their original counterparts.

DGNN: Decoupled Graph Neural Networks with Structural Consistency between Attribute and Graph Embedding Representations

Graph neural networks (GNNs) demonstrate a robust capability for representation learning on graphs with complex structures, showcasing superior performance in various applications. The majority of existing GNNs employ a graph convolution operation by using both attribute and structure information through coupled learning. In essence, GNNs, from an optimization perspective, seek to learn a consensus and compromise embedding representation that balances attribute and graph information, selectively exploring and retaining valid information. To obtain a more comprehensive embedding representation of nodes, a novel GNNs framework, dubbed Decoupled Graph Neural Networks (DGNN), is introduced. DGNN explores distinctive embedding representations from the attribute and graph spaces by decoupled terms. Considering that semantic graph, constructed from attribute feature space, consists of different node connection information and provides enhancement for the topological graph, both topological and semantic graphs are combined for the embedding representation learning. Further, structural consistency among attribute embedding and graph embeddings is promoted to effectively remove redundant information and establish soft connection. This involves promoting factor sharing for adjacency reconstruction matrices, facilitating the exploration of a consensus and high-level correlation. Finally, a more powerful and complete representation is achieved through the concatenation of these embeddings. Experimental results conducted on several graph benchmark datasets verify its superiority in node classification task.

SpecSTG: A Fast Spectral Diffusion Framework for Probabilistic Spatio-Temporal Traffic Forecasting

Traffic forecasting, a crucial application of spatio-temporal graph (STG) learning, has traditionally relied on deterministic models for accurate point estimations. Yet, these models fall short of identifying latent risks of unexpected volatility in future observations. To address this gap, probabilistic methods, especially variants of diffusion models, have emerged as uncertainty-aware solutions. However, existing diffusion methods typically focus on generating separate future time series for individual sensors in the traffic network, resulting in insufficient involvement of spatial network characteristics in the probabilistic learning process. To better leverage spatial dependencies and systematic patterns inherent in traffic data, we propose SpecSTG, a novel spectral diffusion framework. Our method generates the Fourier representation of future time series, transforming the learning process into the spectral domain enriched with spatial information. Additionally, our approach incorporates a fast spectral graph convolution designed for Fourier input, alleviating the computational burden associated with existing models. Numerical experiments show that SpecSTG achieves outstanding performance with traffic flow and traffic speed datasets compared to state-of-the-art baselines. The source code for SpecSTG is available at https://anonymous.4open.science/r/SpecSTG.