University of Cambridge > > Computer Laboratory NetOS Group Talklets > Predicting Mobile User Location through Nonlinear Time Series Analysis

Predicting Mobile User Location through Nonlinear Time Series Analysis

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Stephen Kell.

Accurate and fine-grained prediction of user location and their geographical profile has interesting applications including targeted content and advertisement dissemination, recreational social network tools and many more in areas such as psychology and anthropology. Existing techniques based on linear and probabilistic models are not able to provide accurate prediction of the movement patterns from a spatio-temporal perspective, since they cannot capture the nonlinear characteristics of the behavior of the users if present.

A contribution of this paper is the identification of some degree of determinism, previously uncaptured, in patterns of visits of humans to specific places, at least for the scenarios taken into consideration in this work, based on user GPS position datasets and base station registration data. We then illustrate an approach to location prediction based on nonlinear time series analysis of the arrival and residence times of users in relevant places that are automatically extracted by mining their movement patterns. Moreover, we report about our evaluation over these datasets which confirms a prediction accuracy which ranges between 65% and 90% even after a number of hours. We compare our forecasting results to those obtained by means of the prediction techniques proposed in the literature, showing we have more stable accuracy over time. We also report the performance of an application for dissemination of contents that are characterized by spatio-temporal constraints.

This talk is part of the Computer Laboratory NetOS Group Talklets series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.


© 2006-2021, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity