1. 1 Jong Hee Kang, William Welbourne, Benjamin Stewart, Gaetano Borriello, October 2004, Proceedings of the 2nd ACM international workshop on Wireless mobile applications and services on WLAN hotspots 2005. 04. 26 Database Lab. M.S.1 Kim Ji-Young

2. 2 Introduction Related work Extracting places Experimental evaluation Conclusions and future work Outline

3. 3  Locations  Values measured by the underlying location sensing technologies  Expressed in coordinates or landmarks  e.g., coordinates : (-122.124511, 47.653932)  landmarks : nearby cell tower 34  Places  Locales with semantic meanings to individual users  e.g., home, office, coffee shop Location vs. Place

4. 4  Places are more useful to applications than locations  Locations need to be translated into places Translating Locations into Place(1/2) Location Sensing Technology (-122.124511, 47.653932) My Office Location Place Application Place Translator

5. 5  The ‘Place Translator’ needs mapping information between locations and places  Goal is to generate the mapping information automatically from the user’s behavior (trace of locations) Translating Locations into Place(2/2)

6. 6  Collected using Place Lab, a coordinate-based location system using a database of locations of WiFi hotspots  e.g. User initially stays at place A, then moves to place B and stays there  Important places are those where the user spends a significant amount of time and/or visits frequently Trace of Locations(1/2) AB

7. 7 ImportantPlaces Trace of Locations(2/2)

8. 8 Applying Clustering Algorithms

9. 9  Algorithms require the number of clusters as a parameter  Clusters include unimportant locations (intermediate and transitory locations between truly significant places)  Algorithms require a significant amount of computation Problems Existing Clustering Algorithms

10. 10  Clustering locations along the time axis  A new location is compared with previous locations  If the new location is moving away, starts a new cluster  Then, ignore the clusters with short time duration Time-based Clustering(1/2) Important Places

11. 11 Time-based Clustering(1/2) Time-based Clustering(2/2) Figure2. Existing ClusteringFigure4. Time-based Clustering

12. 12 Two Parameters  Distance threshold (d )  Determines the size of clusters  Time threshold (t )  Determines the number of significant places t d

13. 13 Determining d and t (1/2)

14. 14 Determining d and t (2/2)

15. 15 Experimental Evaluation-Campus Trace

16. 16 Experimental Evaluation-City Scale Trace

17. 17  Our approach can automatically find significant places from the trace of locations  Future work  Connecting to the destination prediction system  To predict a user’s destination from their current location and past observations of their movements (Record the arrival and leaving time to and from the extracted places) Conclusion and Future Work