Share on Facebook Tweet on Twitter Share on LinkedIn Share by email
Nicholas Jing Yuan's Recent Work
  • Functional Regions of a City

Modern cities usually have areas with different social-economic functions, e.g., commercial area, residential area and educational area, either politically driven or self-formed. Identifying the functional regions of a city is critical for many emerging applications, such as social recommendation, computational advertisement and urban computing. In this work, we provide a framework to discover the functional regions of a city, leveraging the data sources from both the human mobility and points of interests (POI).

Human Mobility: With the popularity of geo-enabled mobile devices, many kinds of human mobility data are already available these days. The figure below plots the density scatter of taxi pick-up/drop-off points in the city of Beijing, during a period of 3 months.

Points of Interests: The following figures show the geo-spatial distribution of pub/bar/ktvs (merged as one category) and theaters changing during the past 6 years, in the city of Beijing.

 

 

For the detailed approach, see

Jing Yuan, Yu Zheng, and Xing Xie. Discovering Regions of Different Functions in a City Using Human Mobility and POIs, ACM SIGKDD, 2012.

  • T-Finder: A Recommender System for Finding Passengers and Vacant Taxis.

Have you ever suffered from waiting a long time for a taxicab? Actually, taxi drivers are also upset when cruising on road surfaces for finding passengers. By mining the historical GPS trajectories of taxicabs, T-Finder provides people with nearby locations where they can easily find vacant taxis, and recommends the taxi drivers with some locations, towards which they are most likely to maximize their profit, as well as find potential passengers. This work was presented in

Jing Yuan, Yu Zheng, Liuhang Zhang, Xing Xie and Guangzhong Sun, Where to Find My Next Passenger? Ubicomp 2011

t-finder app

Recently, we extend this work by 1) further analyzing the waiting time for a passenger given his/her current time and location, based on a non-homogeneous Poisson process and 2) estimating the queue length at the locations where taxicabs are parked to wait for passengers. See the details in the following paper:

Nicholas Jing Yuan, Yu Zheng, Liuhang Zhang and Xing Xie, T-Finder: A Recommender System for Finding Passengers and Vacant Taxis, accepted by IEEE Transactions on Knowledge and Data Engineering (TKDE).

  • Morphological Map Segmentation

We demonstrate the detailed approach for segmenting an urban road network into regions bounded by the major roads in the following technical report. This approach was employed in many of our previous papers, such as [1][2] and [3].

Nicholas Jing Yuan, Yu Zheng and Xing Xie, Segmentation of Urban Areas Using Road Networks, MSR Technical Report-2012-65, 2012.

 

[1] Wei Liu, Yu Zheng, Sanjay Chawla, Jing Yuan, and Xing Xie. Discovering Spatio-Temporal Causal Interactions in Traffic Data Streams, Association for Computing Machinery, Inc., 24 August 2011.

[2] Yu Zheng, Jing Yuan, and Xing Xie. Urban Computing with Taxicabs, ACM, 16 September 2011.

[3] Jing Yuan, Yu Zheng, and Xing Xie. Discovering Regions of Different Functions in a City Using Human Mobility and POIs, ACM SIGKDD, 2012.