Multi-Graph Fusion Networks for Urban Region Embedding (IJCAI-22)

This is the implementation of Multi-Graph Fusion Networks for Urban Region Embedding (MGFN) in the following paper:

Shangbin Wu#, Xu Yan#, Xiaoliang Fan*, Shirui Pan, Shichao Zhu, Chuanpan Zheng, Ming Cheng, Cheng Wang, Multi-Graph Fusion Networks for Urban Region Embedding, International Joint Conference on Artificial Intelligence (IJCAI-22), July 23-29, 2022 Messe Wien, Vienna, Austria.[Acceptance rate=15%]

Multi-Graph Fusion Networks for Urban Region Embedding (https://www.ijcai.org/proceedings/2022/0321.pdf) was accepted by IJCAI-2022.

Data
Requirements
QuickStart
These Features are Coming Soon
Citation
Contacts
Reference

Data

Here we provide the processed data. And the Raw Data can be found in: NYC OpenData: https://opendata.cityofnewyork.us/.
We followed the settings in [Zhang et al., 2020] that apply taxi trip data as human mobility data and take the crime count, check-in count, land usage type as prediction tasks, respectively.
For New York City zoning dataset, you can contact with Xu Yan(yanxu97@stu.xmu.edu.cn) or Shangbin Wu(shangbin@stu.xmu.edu.cn).

Requirements

Python 3.7.9,
pytorch 1.5.1,
numpy 1.19.2,
pandas 0.25.3,
sklearn 0.24.1,
geopandas 0.13.0,
shapely 2.0.1

QuickStart

run the command below to train the MGFN:

python mgfn.py

These Features are Coming Soon

The code about...

Visualization of mobility pattern
Generalization ability analysis
Data preprocessing

Citation

Please cite our paper in your publications if this code helps your research.

@article{wu2022multi_graph,
  title={Multi-Graph Fusion Networks for Urban Region Embedding},
  author={Wu, Shangbin and Yan, Xu and Fan, Xiaoliang and Pan, Shirui and Zhu, Shichao and Zheng, Chuanpan and Cheng, Ming and Wang, Cheng},
  journal={arXiv preprint arXiv:2201.09760},
  year={2022}
}

Contacts

Shangbin Wu, shangbin@stu.xmu.edu.cn

Xiaoliang Fan (corresponding author), fanxiaoliang@xmu.edu.cn, https://xiaoliangfan.github.io

Reference

[Zhang et al., 2020] Mingyang Zhang, Tong Li, Yong Li, and Pan Hui. Multi-view joint graph representation learning for urban region embedding. In Christian Bessiere, ed- itor, Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, IJCAI-20, pages 4431–4437. International Joint Conferences on Artificial Intelligence Organization, 7 2020. Special track on AI for CompSust and Human well-being.

wushangbin/MGFN