### Urban Mobility Pattern Mining Based on Regional Dependencies

 Mobility pattern mining plays an important role in city planning and traffic infrastructure construction. In this work, we focus on mobility patterns that reflect the crowd dynamics between city regions to help urban district planning. We formulate the problem by clustering origin and destination points in two views to maximize correlations between origin and destination regions. This formulation can extract complex mobility patterns from many origin regions to many destination regions without region overlapping, which can be easily visualized and interpreted. To efficiently extract such patterns from taxi trip data, we proposed kernelized ACE, a novel co-clustering algorithm based on information theory. Our method is tested on both synthesized and real world data. For synthesized data, it perfectly captures the patterns we generated. For Beijing and New York City data, it shows region partitions coincide with the topology and mobility patterns in each city.

#### Publication

##### Conference Version
 Jing Lian, Yang Li, Weixi Gu, Shao-Lun Huang, Lin Zhang. Joint Mobility Pattern Mining with Urban Region Partitions, 15th EAI International Conference on Mobile and Ubiquitous Systems (MobiQuitous). November 5-7, 2018 pdf
Bibtex
@inproceedings{lian2018joint,
title={Joint Mobility Pattern Mining with Urban Region Partitions},
author={Lian, Jing and Li, Yang and Gu, Weixi and Huang, Shao-Lun and Zhang, Lin},
booktitle={Proceedings of the 15th EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services},
pages={362--371},
year={2018},
organization={ACM}
}

##### Journal Version

This work is an extension to our conference paper with more detailed literature review, thorough feature and cluster results evaluation, new case studies using POI information of NYC and a multi-year mobility analysis of Beijing.

 Jing Lian, Yang Li, Weixi Gu, Shao-Lun Huang, Lin Zhang. Mining Regional Mobility Patterns for Urban Dynamic Analytics. Mobile Network Applications, Vol 25, 459–473, 2020. pdf
Bibtex
@article{lian2020mining,
author = {Lian, Jing and Li, Yang and Gu, Weixi and Huang, Shao-Lun and Zhang, Lin},
year = {2020},
month = {04},
title = {Mining Regional Mobility Patterns for Urban Dynamic Analytics},
volume = {25},
journal = {Mobile Networks and Applications}
}