Traffic prediction serves as a cornerstone for systems and network services such as the Web of Vehicles (WoV), online navigation, and smart city applications. Despite the proliferation of model architectures in recent years, existing approaches often suffer from highly customized structures and weak transferability, making it difficult to cope with increasing task heterogeneity and modeling complexity. To address these challenges, we propose ST-LEGO, a modular assembly framework driven by large language models (LLMs) that supports flexible structural composition and automated code generation. ST-LEGO employs a multi-agent collaborative system comprising a Prompt Agent, Assemble Agent, and Code Agent, which are responsible for understanding task requirements, dynamically assembling structural modules, and automatically generating executable PyTorch code. By introducing a standardized module library and an intermediate structural description language (DSL), the framework enables controllable generation, reusable composition, and cross-task generalization of model architectures. Empirical results on multiple real-world traffic datasets demonstrate that models generated by ST-LEGO achieve superior accuracy, structural diversity, and convergence compared to a wide range of manually designed baselines. These results highlight the unique potential and scalability of LLMs as structural architects for traffic prediction, offering a new paradigm for integrating language models into web-interactive intelligent transportation systems.

Fine-grained traffic management and prediction are fundamental to key applications such as autonomous driving, lane change guidance, and traffic signal control. However, obtaining lane-level traffic data has become a critical bottleneck for data-driven models due to limitations in the types and number of sensors and issues with the accuracy of tracking algorithms. To address this, we propose the Fine-grained Road Traffic Inference (FRTI) task, which aims to generate more detailed lane-level traffic information using limited road data, providing a more energy-efficient and cost-effective solution for precise traffic management. This task is abstracted as the first scene of the spatio-temporal graph node generation problem. We designed a two-stage framework-RoadDiff-to solve the FRTI task. This framework leverages the Road-Lane Correlation Autoencoder-Decoder and the Lane Diffusion Module to fully utilize the limited spatio-temporal dependencies and distribution relationships of road data to accurately infer fine-grained lane traffic states. Based on existing research, we designed several baseline models with the potential to solve the FRTI task and conducted extensive experiments on six datasets representing different road conditions to validate the effectiveness of the RoadDiff model in addressing the FRTI task. The relevant datasets and code are available at https://github.com/ShuhaoLii/RoadDiff.

Traffic prediction has long been a focal and pivotal area in research, witnessing both significant strides from city-level to road-level predictions in recent years. With the advancement of Vehicle-to-Everything (V2X) technologies, autonomous driving, and large-scale models in the traffic domain, lane-level traffic prediction has emerged as an indispensable direction. However, further progress in this field is hindered by the absence of comprehensive and unified evaluation standards, coupled with limited public availability of data and code. In this paper, we present the first systematic classification framework for lane-level traffic prediction, offering a structured taxonomy and analysis of existing methods. We construct three representative datasets from two real-world road networks, covering both regular and irregular lane configurations, and make them publicly available to support future research. We further establishes a unified spatial topology structure and prediction task formulation, and proposes a simple yet effective baseline model, GraphMLP, based on graph structure and MLP networks. This unified framework enables consistent evaluation across datasets and modeling paradigms. We also reproduce previously unavailable code from existing studies and conduct extensive experiments to assess a range of models in terms of accuracy, efficiency, and applicability, providing the first benchmark that jointly considers predictive performance and training cost for lane-level traffic scenarios. All datasets and code are released athttps://github.com/ShuhaoLii/LaneLevel-Traffic-Benchmark.

Lane-level traffic prediction is crucial for refined smart city applications, yet the scarcity and quality issues of datasets hinder its development. To overcome these challenges, this study introduces a novel M ulti- c hannel g raph-structured V ariational A uto E ncoder model, McgVAE. This model integrates road-level information to provide a global perspective for lane prediction and performs integrated tasks through three interconnected channels: the road-level channel ensures accurate prediction of road traffic states and communicates closely with the data quality channel to share historical and predicted road information; the data quality channel leverages road-level information to identify and correct missing and noisy data; and finally, the lane channel uses the aforementioned information for lane-level traffic prediction. After extensive experimental comparisons with multiple baseline models across three datasets, the McgVAE model demonstrated outstanding predictive performance and the ability to handle data missingness and noise

Numerous algorithms have been proposed for discovering denial constraints (DCs), which are essential and effective for maintaining data consistency. However, existing methods only focus on discovering the complete set of DCs, often resulting in hundreds or even tens of thousands of discovered rules. Such a large number of DCs are impractical for users to verify and utilize. Besides, these methods overlook the intent of users, which requires the discovered DCs to be succinct, relevant, and diverse concurrently. To address these limitations, we introduce DCMiner, a deep reinforcement learning (DRL)-based framework that produces rules satisfying user preferences. Specifically, we first model the discovering process via a kCover Markov decision process to improve efficiency. Then, a graphQ model is introduced to capture the data distribution and facilitate the discovery of DCs. Lastly, we design a reward function that flexibly integrates both objective and subjective criteria to align the discovered rules with user intent, and we propose an efficient training process. Extensive experiments on both real-world and synthetic datasets show that DCMiner can discover succinct, relevant, and diverse rules.

With the widespread application of intelligent transportation systems and navigation software, traffic prediction should be modeled in finer granularity to facilitate lane-changing guidance and congestion mitigation. However, existing studies divide the road network into continuous segments which assumes different lanes share the same spatio-temporal patterns. This paper proposes a novel lightweight, attention-based, fully convolutional model, named the Spatio-Temporal Attention- Based Convolutional network (ST-ABC), where lane segments are treated as graph nodes and dynamically models the adjacent spatial dependencies using local attention graph convolution. The attention-based dilated convolutions can process longer sequence periods in parallel, and a global attention layer allows individual nodes to be associated with the global context. By setting a target window, it can further reduce unnecessary computations and improve the prediction effect for the targeted area. Further-more, the ST-ABC model facilitates the simultaneous integration of spatio-temporal information and relational distance metrics among lane segments, enriching the granularity of multi-scaled spatial prediction. Empirical evaluations conducted on two real-world datasets substantiate the augmented efficacy of the STABC model in comparison to established models, with a marked prominence in long-term prediction scenarios.

Order dependencies (ODs) state ordering specifications between attributes, and have been proven effective in query optimization for sorting operations. In this paper we investigate the problem of set-based OD discovery, for automatically finding hidden ODs from data. We tackle the problem with a novel level-wise hybrid strategy. With a given relational instance r, we discover ODs from a sample (subset) of r, validate the discovered ODs on r and refine the sample by leveraging the validation, in a level-by-level manner according to the lattice of set-based ODs. This process continues until the discovery result on the sample converges to that on r. We prove that a dynamic sample whose size keeps growing can be used in the process without affecting the correctness and completeness of the discovery result, and present techniques to incrementally refine the sample on demand. We also enhance our method with multi-threaded parallelism. On a host of datasets, our method is faster than the state-of-the-art method up to orders of magnitude even when the parallelism of our approach is disabled, and achieves up to a 4.5x self-relative parallel speedup with 6 threads.

The pervasiveness of GPS-enabled devices and wireless communication technologies results in a proliferation of traffic data in intelligent transportation systems, where traffic prediction is often essential to enable reliability and safety. Many recent studies target traffic prediction using deep learning techniques. They model spatio-temporal dependencies among traffic states by deep learning and achieve good overall performance. However, existing studies ignore the bias on traffic prediction models, which refers to non-uniformed performance distribution across road segments, especially the significantly poor prediction results on certain road segments. To solve this issue, we propose a framework named spatio-temporal mixture-of-experts (ST-MoE) that aims to eliminate the bias on traffic prediction. In general, we refer to any traffic prediction model as the based model, and adopt the proposed ST-MoE framework as a plug-in to debias. ST-MoE uses stacked convolution-based networks to learn spatio-temporal representations of individual patterns of road segments and then adaptively assigns appropriate expert layers (sub-networks) to different patterns through a spatio-temporal gating network. To this end, the patterns can be distinguished, and biased performance among road segments can be eliminated by experts tailored for specific patterns, which also further improves the overall prediction accuracy of the base model. Extensive experimental results on various base models and real-world datasets prove the effectiveness of ST-MoE.

This work introduces a pioneering, engineerable approach to 3D visual localization(3DVG). Current challenges for 2D Visual Grounding (2DVG) and 3DVG are summarized: Absence of Depth Information in 2DVG, Memory and Computational Demands of Global Point Clouds, Limitations in Dynamic Scenarios, and Limited Understanding of Spatial Localization Reference Frames. Our solution proposes a Re_3DVG method for fragmented point cloud scenarios. Utilizing instance segmentation and transformer models, our approach offers a potent mechanism for establishing robust correspondences between text queries and object instances within the shared visible range. The introduction of a FragCloud3DRef dataset, grounded in ScanNet and supplemented with RGB data, object segmentation, and textual descriptions, fortifies the effectiveness of our proposed model. Experimental outcomes display that our model excels beyond conventional 3DVG and 2DVG models, establishing a formidable benchmark for future research within this discipline. The code source and dataset are open at https://github.com/zehantan6970/Reimagining_3DVG.

Lexicographical order dependencies state relationships of order between lists of attributes. They naturally model the order-by clauses in SQL queries, and are proven useful in query optimizations concerning sorting. Despite their importance, order dependencies on a dataset are typically unknown and are too costly, if not impossible, to design or discover manually. Techniques for automatic order dependency discovery are recently studied. It is challenging for order dependency discovery to scale well, since it is by nature factorial in the number$m$of attributes and quadratic in the number$n$of tuples. In this article, we adopt a strategy that decouples the impact of$m$from that of$n$, and that still finds all minimal and valid lexicographical order dependencies. We present carefully designed data structures, a host of algorithms and optimizations, and an enhanced strategy combined with multithreaded parallelism, for an efficient implementation. Using a host of real-life and synthetic datasets, we experimentally verify our approach is up to orders of magnitude faster than the state-of-the-art methods, and can deliver better results with an improved definition of minimal attribute lists.

Question retrieval aims to increase the accessibility of the community Question Answer (cQA) archives and has attracted increasing research interests recently. In this paper, we present a novel method for improving the question retrieval performance by investigating the question term selection and weighting as well as reranking results. Different from previous work, we propose a hierarchical question classification method with a sparse regularization to mimc user's question labeling in cQAs. Based on the hierarchical classification, we explore the local context of the question for term selection and reranking results and then integrating them into our proposed general question retrieval framework. The experimental results on a Yahoo! Answers dataset show the effectiveness of our method as compared to existing general question retrieval models and some state-of-the-art methods of utilizing category information for question retrieval.

We present a hierarchical kernelized classification model for the automatic classification of general questions into their corresponding topic categories in community Question Answering service (cQAs). This could save many efforts of manual classification and facilitate browsing as well as better retrieving of questions from the cQA archives. To deal with the challenge of short text message of questions, we explore and optimally combine various cQA features by introducing multiple kernel learning strategy into the hierarchical classification framework. We propose a hybrid regularization approach of combining orthogonal constraint and L1 sparseness in our framework to promote the discriminative power on similar topics as well as sparsing the model parameters. The experimental results on a real world dataset from Yahoo! Answers demonstrate the effectiveness of our proposed model as compared to the state-of-the-art methods and strong baselines.