Web-Enhanced GPS

  • Authors:

  • Ramaswamy Hariharan;John Krumm;Eric Horvitz

  • Affiliations:

  • School of Information and Computer Sciences, University of California, Irvine, CA;Microsoft Research, Microsoft Corporation, One Microsoft Way, Redmond, WA;Microsoft Research, Microsoft Corporation, One Microsoft Way, Redmond, WA

  • Venue:
  • LoCA'05 Proceedings of the First international conference on Location- and Context-Awareness
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Location-based services like reminders, electronic graffiti, and tourist guides normally require a custom, location-sensitive database that must be custom-tailored for the application at hand. This deployment cost reduces the initial appeal of such services. However, there is much location-tagged data already available on the Web which can be easily used to create compelling location-aware applications with almost no deployment cost. Such tagged data can be used directly in applications as well as to provide evidence in models of activity. We describe three applications that take advantage of existing Web data combined with location measurements from a GPS receiver. The first application, “Pinpoint Search”, finds web pages of nearby places based on GPS coordinates, queries from a Web mapping service, and general Web searches. The second application, “XRay”, uses the mapping service to find businesses in a building by pointing a GPS-equipped electronic compass at the building. The third application is called “Travelogue”, and it builds a map and clickable points of interest to help automatically annotate a trip based on GPS coordinates. Finally, we discuss the use of Web-based data as rich sources of evidence for probabilistic models of a user's activity, including a means for interpreting the explanation for the loss of Web signals as users enter structures.