Recognizing user context is important in making mobile devices simple to use. The device and the underlying mobile service can provide a personalized user interface that adapts to the usage situation. The device can infer parts of the context of the user from features extracted from on-board sensors that make measurements of acceleration, noise level, luminosity, humidity, etc. In this paper we justify why generating different level descriptions of user context is necessary for the development of useful applications. We also show how it is possible to generate higher level contexts. A practical method based on the clustering of symbol string data using the SCM algorithm is described. This is applied to the analysis of a real data set from sensor recordings. Based on this data and using the various ideas presented, different levels of user context are generated and a simple method for changing a mobile device profile are described.
|Publication status||Published - 2003|
|MoE publication type||Not Eligible|
|Event||Workshop on Artificial Intelligence in Mobile System - Seattle, United States|
Duration: 12 Oct 2003 → 12 Oct 2003
|Conference||Workshop on Artificial Intelligence in Mobile System|
|Period||12/10/03 → 12/10/03|
Flanagan, J. A., Himberg, J., & Mäntyjärvi, J. (2003). A hierarchical approach to learning context and facilitating user interaction in mobile devices. Paper presented at Workshop on Artificial Intelligence in Mobile System, Seattle, United States.