Learning Significant Locations and Predicting User Movement with GPS

Ashbrook, D. 2002. Learning Significant Locations and Predicting User Movement with GPS. In Proceedings of the 6th IEEE international Symposium on Wearable Computers (October 07 – 10, 2002). ISWC. IEEE Computer Society, Washington, DC, 101.

This paper investigates the creation of a predictive model of the user’s future movements, based on the clustering of GPS data and their incorporation into a Markov model that can be consulted for use with a variety of location-aware applications.

The contribution of this authors lies in the determination of users’ significant places (clustering) and predicting the next move (Markov model). One limitation of this approach is that changes in schedule may take a long time to be reflected in the model.

They mention problems in using GPS:

While in many respects GPS is an ideal sensor, some problems were encountered. Although Selective Availability has been turned off, the accuracy of our GPS receiver was 15 meters; this means that the same physical location will have a different GPS coordinate from day to day.

Relation to my thesis: In their litterature review, they mention some methods for inferring location based on GPS traces (start and end points of trips, building detection) that are close to the issues of a project I am involved in:

In their investigations of automatic travel diaries [16], Wolf et. al. used stopping time to mark the starting and ending points of trips. In their work on the comMotion system [10], Marmasse and Schmandt used loss of GPS signals to detect buildings. When the GPS signal was lost and then later re–acquired within a certain radius, comMotion considered this to be indicative of a building. This approach avoided false detection of buildings when passing through urban canyons or suffering from hardware issues such as battery loss.

[10] Natalia Marmasse and Chris Schmandt. Location–aware information delivery with ComMotion. In HUC, pages 157–171, 2000.

[16] Jean Wolf, Randall Guensler, and William Bachman. Elimination of the travel diary: An experiment to derive trip purpose from GPS travel data. In Notes from Transportation Research Board, 80th annual meeting, Washington, D.C., 2001.

This paper does not mention the many issues the prediction and inferences could create on the user level. However, their method could be interesting for post-game analysis of users intentions in CatchBob!