Urban Computing with Taxicabs

1. Urban Computing with
Taxicabs
UbiComp '11 Proceedings of the 13th international
conference on Ubiquitous computing
Yu Zheng, Yanchi Liu, Jing Yuan, Xing Xie
Microsoft Research Asia, Beijing
University of Science and Tech Beijing

2. • Flawed(less effective) urban planning
• GPS trajectories of taxicabs
Traveling in urban areas.
• To detect Salient traffic problem
• Result can evaluate the effectiveness of the carried
out planning
•newly built road, subway line.

3. • Urban computing for urban planning is one of the
most significant application scenarios in the urban
spaces.
• Kindberg, T., Chalmers, M., Paulos, E. Gest editors’
introduction: Urban computing. Pervasive computing. 6,
3(2007), 18-20.
• Shklovski, I., Chang, M. F. Urban Computing-Navigating
Space and Context. IEEE Computer Society. 39, 9 (2006),
36.

4. Method
• Partition a city map into disjoint regions using
major roads as boundaries.
• The region pairs frequently detected across many
days will be regarded as the flawed planning.

6. Temporal partition
• Workday and restday
• Time slot

7. Contribution
• Traffic modeling: matrix of regions
• Each item consists of a set features representing the
connection between two regions.
• Flaw detection: flawed region pairs(skyline) from
matrix of each day using a skyline operator.

8. Qualitative analysis
• The differential and the identical between 2009
and 2010.
• Traffic
• Volume
• Speed
• Infer phenomenon from computational results
• The traffic conditions of Beijing become worse in
2010(compared with 2009).
• A few people traveling by taxis before bought their own
private cars in 2010.

9. My conclusion
• In my next paper, I also will focus on traffic
problem for urban planning.
• The way it partition map is not good.
• No comparison
• Qualitative analysis are not strict.

10. My conclusion
• In my next paper, I also will focus on traffic
problem for urban planning.
• The way it partition map is not good.
• No comparison
• Qualitative analysis are not strict.