XTraffic: A Dataset Where Traffic Meets Incidents with Explainability and More

Authors

Xiaochuan Gou*, Ziyue Li*, Tian Lan, Junpeng Lin, Zhishuai Li, Bingyu Zhao, Chen Zhang, Di Wang, Xiangliang Zhang

*The authors contributed equally to this work.

Introduction

Welcome to the XTraffic dataset repository! This dataset integrates traffic and incident data across a large-scale region, covering 16,972 traffic nodes over the entire year of 2023. XTraffic includes time-series data on traffic flow, lane occupancy, and average vehicle speed, along with spatiotemporally-aligned incident records across seven different classes. Each node also features detailed physical and policy-level meta-attributes of lanes. Our goal is to enhance the interpretability and practical applications of traffic management and safety analysis through this comprehensive dataset.

How to Use

Clone the Repository:

git clone https://github.com/your-username/XTraffic-dataset.git
cd XTraffic-dataset

Install Dependencies: Ensure you have the necessary dependencies installed. You can use the requirements.txt file to set up your environment.
```
pip install -r requirements.txt
```
Download the Dataset: The dataset is available for download at the provided URLs. Follow the links to download the required files.
Run Examples: Example scripts are provided to demonstrate how to use the dataset for various tasks like incident classification, and causal analysis.

For traffic forecasting:

We recommend you use the repository LargeST for traffic forecasting tasks.

For incident classification:

Refer to readme.md in the classification floder.

For local&global causal analysis:

Refer to readme.md in the causal_analysis folder.

Meta Data

The meta data for the sensors and incidents can be accessed through the following URL: Meta Data URL

Feel free to explore and contribute to the repository. If you have any questions or suggestions, please open an issue or contact the authors.

Citation

If you use our dataset in your research, please cite our paper:


@misc{gou2024xtrafficdatasettrafficmeets,
      title={XTraffic: A Dataset Where Traffic Meets Incidents with Explainability and More}, 
      author={Xiaochuan Gou and Ziyue Li and Tian Lan and Junpeng Lin and Zhishuai Li and Bingyu Zhao and Chen Zhang and Di Wang and Xiangliang Zhang},
      year={2024},
      eprint={2407.11477},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2407.11477}, 
}