A Brief Journey through Foursquare User Check-ins in Cardiff

Informal talk.
Venue: FTS Seminar, Cardiff University School of Computer Science & Informatics.

Transcript

1. A brief journey through Foursquare
   user check-ins in Cardiff
   FTS
   6/April/2011
   Matt Williams
   (joint project with Martin Chorley)
   

   View Slide

2. Talk overview
   1. Foursquare 2. Data collection
   3. Results &
   visualisation
   4. Tools
   

   View Slide

3. Foursquare
   • A “location-based online social network”
   • Founded in 2007 in New York
   • Users ‘check-in’ to their current venue
   • Venues are user-contributed
   • ‘Mayorship’ of a venue given to the user who
   has visited the most in two months
   • Points awarded for checking in
   

   View Slide

4. World Foursquare usage
   Text
   http://blog.foursquare.com/2011/01/24/2010infographic/
   

   View Slide

5. Why use Foursquare?
   • To meet up with friends?
   • For deals at your favourite venues?
   • ‘Life-logging’?
   • Embedded social gaming?
   

   View Slide

6. Why use Foursquare?
   • Answer: Science!
   • A rich dataset:
   • Human mobility
   • Geocoded venues
   • Social graph
   

   View Slide

7. Collection method
   • Problem: user check-in histories are not public
   (understandably!)
   • ...however, venues do have a “currently here” list
   that users can opt-in to publicly appear on
   • Solution:
   1. collect a set of venues
   2. monitor the venues in real time!
   

   View Slide

8. Data collection
   • Collected 4,314 venues in a
   9km x 7km region
   • Restricted monitoring to
   about 2,000 venues
   • Began monitoring on
   Monday 21st March
   

   View Slide

9. Dataset summary
   • Collected Mon 21st March - Sat 2nd April
   (13 days)
   • 3,394 check-ins
   • 767 venues visited
   • 4.4 check-ins per venue
   • 716 users with check-ins
   • Friends per user: average 126, median 19
   

   View Slide

10. When do people check in?
    

    View Slide

11. Check-ins over time
    

    View Slide

12. How does the popularity of venues vary?
    

    View Slide

13. Top 20 venues
    

    View Slide

14. Venue popularity
    linear-linear plot
    

    View Slide

15. Venue popularity
    log-log plot
    

    View Slide

16. Where in Cardiff do check-ins occur?
    

    View Slide

17. Check-in heatmap
    

    View Slide

18. How does check-in activity vary between
    users?
    

    View Slide

19. Top 20 users
    

    View Slide

20. User check-ins
    

    View Slide

21. How many friends does a user have?
    

    View Slide

22. Friendships
    

    View Slide

23. Friendships among Cardiff visitors
    

    View Slide

24. [An animation of checkins!]
    

    View Slide

25. Tools
    • Python (important! :))
    • matplotlib
    • Microsoft Excel
    • Processing
    • Google Earth
    • ffmpeg
    • heatmap.py (by jjguy)
    • www.getlatlon.com
    

    View Slide

26. Thanks!
    Any questions?
    Twitter & 4sq: @voxmjw
    

    View Slide