Using web-sources for location based systems on mobile phones

The paper describes the implementation of a context aware location and time based system on a normal mobile phone, with a minimum of requirements (Bluetooth, Internet access, Java ME, and a simple webbrowser). We can answer the following queries based on data crawled from the web and enriched with location coordinates:
• Where are my friends?
• What movies are playing, right here, right now?
• I need to park my car. Where is the closest parking lot with free space?
• What is a good place to have dinner around here?
• What is the history of this building?
• What experiences have other people had at this location?
Similar systems have been built, but all in a limiting setting (specific hardware/software, specially built devices, limited areas). While many systems only use their own data, we use public sources from the Internet that can be coupled to location. This paper shows that it is possible and in fact rather easy to build useful location based systems on the kind of device which needs it most: your own mobile phone.
The challenge for location based services is thus located in obtaining valuable location and time based information from the web. We used 7 different web-sources, each with their own data format. These use-cases indicate systembottlenecks and yield a useful set of requirements for emerging standards on location based data. A full version of this paper is available at http://www.science.uva.nl/~marx/pub/nulaz.pdf.