1. 國立雲林科技大學 National Yunlin University of Science and Technology 11 Discovering Personal Gazetteers: An Interactive clustering Approach Changqing Zhou, Dan Frankowski, Pamela Ludford, Shashi Shekhar, and Loren Terveen 2004.GIS Presenter : Jin-Wei Lin Advisor : Professor Chung-Chian Hsu 2007/01/17

2. N.Y.U.S.T. I. M. 22 Outline Motivation Objective Introduction Approach Evaluation Framework Experiment Conclusion Comments

3. N.Y.U.S.T. I. M. 33 Motivation Personal gazetteers record individuals’ most important places. There has been previous work on this problem, e.g., ad hoc algorithms, K-mean clustering, but both approaches have shortcomings.

4. N.Y.U.S.T. I. M. 44 Objective This paper explores the use of novel semi- automatic techniques to discover gazetteers from users’ travel patterns.

5. N.Y.U.S.T. I. M. 55 Introduction Location data storage Location sensing Cluster and visualization

6. N.Y.U.S.T. I. M. 66 Approach- Exploratory Approach The loss of the GPS signal is interpreted as a significant cue, namely that a building has been entered. When the signal has been lost within a given radius on three different occasions the agent infers that this location (building) is interesting.

7. N.Y.U.S.T. I. M. 77 Approach- Clustering Approach Density-based Clustering K-Mean Clustering K-Means Clustering Density-based Clustering DJ-clustering The number of clusters OXX Sensitive to noise OXX DeterministicXOO ShapeCircleArbitrary ParametersclustersEps and MinPts Efficiency---MediumFast DJ-clustering

8. N.Y.U.S.T. I. M. 88 DJ-clustering P q1q1 q2q2 q3q3 Cluster C X X X X Speed greater than 0 Frequently stops Move small distance

9. N.Y.U.S.T. I. M. 99 Evaluation Framework

10. N.Y.U.S.T. I. M. 10 Experiment Dataset K-MeansDJ-cluster manual

11. N.Y.U.S.T. I. M. 11 Conclusions We currently are in the process of analyzing the results, which will provide a useful accuracy baseline for personal gazetteer discovery algorithms. This paper proposes a deterministic, density based clustering algorithm that also uses temporal techniques to reduce the number of uninteresting places that are discovered.

12. N.Y.U.S.T. I. M. 12 Comments Advantage Improve DBSCAN efficiency Drawback … Application Identify personal characteristics