Spatio-temporal Personalised Web Search

With the proliferation of mobile technologies and the advent of GPS enabled personal devices, spatial as well as temporal data related to a user’s presence in different places over time are becoming available. Such data can be useful in understanding and building a personal user profile that reflects their behaviour, activities and general interests. Existing applications, such as Location-based services (LBS) normally rely on 'instantaneous' space and time data values representing user's location at a certain point in time. Several approaches have been proposed in the literature that considers the use of multiple/historical data values of user’s location information collected over time. However, only simple numerical interpretations of such information have been presented. In this work, it is argued that a rich spatio-temporal data model of Place is needed that incorporates, as well as facilitates, the analysis and discovery of place semantics, such as for example, the type of place, services offered by the place and relationships between places. Ultimately, such a data model will allow for the analysis of collected spatio-temporal data logs to support the development of rich personal user profiles.