Traffic forecasting is pivotal but challenging due to intricate spatio-temporal dynamics. Existing models often apply a uniform spatial mechanism across distinct temporal scales and rely on static feature embeddings. Consequently, they are inadequate in capturing scale-specific spatial heterogeneity and dynamic feature interdependencies. To address these limitations, we propose the Multi-Scale Spatio-Temporal Attention Network (MSSTAN) with a novel dual-branch architecture: (1) A Global-Local Feature Attention Network (GLFAN) that explicitly decouples spatial interactions across decomposed temporal components to capture multi-scale spatial patterns; and (2) A Spatio-Temporal Feature Attention Network (STFAN) that dynamically recalibrates feature importance based on specific spatio-temporal contexts. A dynamic branch fusion mechanism integrates these branches to optimally aggregate their complementary views. Extensive experiments on five real-world datasets demonstrate that MSSTAN achieves state-of-the-art or highly competitive performance, validating its efficacy for traffic forecasting.

Traffic flow forecasting task plays an essential role in intelligent transportation systems. Accurately capturing the intricate spatio-temporal dependencies in traffic network signals is the core of precise prediction. Recently, a paradigm that models spatio-temporal dependencies through graph neural networks and time series models has become one of the most promising methods to solve this problem. However, existing methods still have limitations due to ineffectively modeling dynamic spatial dependencies and high time and space complexity. To address these issues, we propose a simplifying and powerful general spatio-temporal traffic flow forecasting model called LightST. Specifically, LightST first embeds temporal covariates and spatial position information to enhance the spatio-temporal modeling capabilities. Then, stacked temporal linear layers are introduced to capture temporal dependencies efficiently. Finally,we propose a concise adaptive spatio-temporal embedding graph convolution method to extract implicit spatial dependencies over time via dynamic graph convolution with adaptive spatio-temporal embedding graph generation. Extensive experiment results on four public traffic flow datasets demonstrate the superiority of our LightST concerning computational efficiency and prediction performance.

Spatio-temporal prediction is crucial in numerous real-world applications, including traffic forecasting and crime prediction, which aim to improve public transportation and safety management. Many state-of-the-art models demonstrate the strong capability of spatio-temporal graph neural networks (STGNN) to capture complex spatio-temporal correlations. However, despite their effectiveness, existing approaches do not adequately address several key challenges. Data quality issues, such as data scarcity and sparsity, lead to data noise and a lack of supervised signals, which significantly limit the performance of STGNN. Although recent STGNN models with contrastive learning aim to address these challenges, most of them use pre-defined augmentation strategies that heavily depend on manual design and cannot be customized for different Spatio-Temporal Graph (STG) scenarios. To tackle these challenges, we propose a new spatio-temporal contrastive learning (CL4ST) framework to encode robust and generalizable STG representations via the STG augmentation paradigm. Specifically, we design the meta view generator to automatically construct node and edge augmentation views for each disentangled spatial and temporal graph in a data-driven manner. The meta view generator employs meta networks with parameterized generative model to customize the augmentations for each input. This personalizes the augmentation strategies for every STG and endows the learning framework with spatio-temporal-aware information. Additionally, we integrate a unified spatio-temporal graph attention network with the proposed meta view generator and two-branch graph contrastive learning paradigms. Extensive experiments demonstrate that our CL4ST significantly improves performance over various state-of-the-art baselines in traffic and crime prediction. Our model implementation is available at the link: https://github.com/HKUDS/CL4ST.

Accurate traffic forecasting, the foundation of intelligent transportation systems (ITS), has never been more significant than nowadays due to the prosperity of smart cities and urban computing. Recently, Graph Neural Network truly outperforms the traditional methods. Nevertheless, the most conventional GNN-based model works well while given a predefined graph structure. And the existing methods of defining the graph structures focus purely on spatial dependencies and ignore the temporal correlation. Besides, the semantics of the static pre-defined graph adjacency applied during the whole training progress is always incomplete, thus overlooking the latent topologies that may fine-tune the model. To tackle these challenges, we propose a new traffic forecasting framework-Spatio-Temporal Latent Graph Structure Learning networks (ST-LGSL). More specifically, the model employs a graph generator based on Multilayer perceptron and K-Nearest Neighbor, which learns the latent graph topological information from the entire data considering both spatial and temporal dynamics. Furthermore, with the initialization of MLP-kNN based on ground-truth adjacency matrix and similarity metric in kNN, ST-LGSL aggregates the topologies focusing on geography and node similarity. Additionally, the generated graphs act as the input of the Spatio-temporal prediction module combined with the Diffusion Graph Convolutions and Gated Temporal Convolutions Networks. Experimental results on two benchmarking datasets in real world demonstrate that ST-LGSL outperforms various types of state-of-art baselines.

Abstract Nowadays, people’s buying or availing services decisions are subject to online available reviews/opinions. The authenticity of these reviews/opinions is dubious, as there exist many fake reviews posted to attain monetary benefits by promoting their own or demoting the competitor’s products or services known as review spam. Although the number of spam is relatively less than that of normal reviews in real-life, this class imbalance is a critical concern in review spam detection. The performance degrades when the classifier skew towards the majority class. Moreover, efficient feature selection is essentially needed for this issue. The purpose of this study is to develop a framework based on different effective feature selection along with data balancing techniques. Validation results show that our proposed framework commendably copes up with the review spam issue and a higher precision on the real-life dataset. Further, we tested the sensitivity of our proposed framework using both parametric and non-parametric tests and found it significant.

Incremental technique is an efficient mechanism for dealing with dynamic knowledge discovery. The fuzzy probabilistic rough set model on dual universes (FPRSMDU) is an integrated generalization of classic rough set theory (RST) on fuzziness, probability and dual universes. Although a significant number of RST based research efforts have been directed toward developing incremental algorithms to speed up computation of approximations, feature selection, as well as rule extraction in the context of dynamical information systems, there remains lack of effort towards incorporating the incremental method into knowledge updating in the framework of FPRSMDU. Approximations of FPRSMDU are fundamental concepts, which can be used for knowledge discovery in big data or other related work, need to be updated effectively when the objects of two universes vary with time. In light of these issues, an incremental approach for updating approximations in FPRSMDU is proposed while multiple objects inserting into or deleting from the two universes. The validity of the proposed method has been exemplified by employing an illustrative example.

The rough set model over dual universes (RSMDU) as a generalized model of classical rough set theory (RST) on the two universes has been well studied with the objective to establishment of model and discussion of its corresponding properties. Approximations of a concept in RSMDU, which may further be applied to knowledge discovery or related work, need to be updated effectively under a dynamic environment. Despite recent advances in using the incremental method to speed up updating approximations of RST, there has been little effort toward incorporating the incremental method into computing approximations under RSMDU. This paper proposes an incremental learning approach for updating approximations in RSMDU when the objects of two universes vary with time. An illustration is employed to show the proposed method. Extensive experimental results on various real and synthetic data sets verify the effectiveness of the proposed incremental updating method while comparing with the nonincremental method.