2003 – Lyon , France LOCATION-BASED SPATIAL MODELLING USING ONTOLOGY

1. INTRODUCTION The development of mobile communication technology has released users from the constraints of desktop-bound activities; however, the type and format of services offered are mostly monotonous and mundane, focusing on email and SMS exchanges. Emerging technologies, such as Bluetooth, UMTS [1], GPRS, W-LAN have made it possible to provide better mobile communication services using a variety of geographical scales and approaches. In particular, significant improvements in terms of data transfer rate, wireless connectivity, data-format transformation, display techniques and power supply, increasingly extend mobile communications from the pure text and audio transmission to mobile multimedia. Beyond the technical breakthrough, new social and economical life patterns appear, as more users get equipped with portable and embeddable devices, e.g. GPS and other positional sensors, PDAs (portable digital assistants), smart watches [2], mobile phones [3], cameras and home appliances. Currently, most services provide access via mobile phones to public data stores for public transport timetables, weather broadcast, correlative market tendency, book literary performance tickets, movies [3], etc. In the near future, the focus is expected to move towards more interactive services, i.e. full support of a range of personalized and location-based services. That will materialize the idea of a mobile office or sophisticated way finding services. An example scenario for the latter is described in sect. 2. In particular, demand for location-based services has already appeared, and prompted research in different fields such as entertainment appliances, real-time environment monitoring (e.g. traffic control), information on the move (e.g. position-enabled tourism), or interactive mobile services (e.g. virtual bus). Relevant work includes, for instance, investigations into using a data warehouse as an information centre to provide location-based services [4]. The data warehouse integrates and distributes information on the basis of the users' particular situation, preferences and needs. The task involves dynamic information generation through the interaction with users, spatial-temporal information extraction from different dimensions and measures. In [5] Pfoser & al. propose an ad-hoc database model to locate correlative data stores and exchange similar information within a specific community. The model is mainly composed of Data Handlers, Data Stores and proxies, and uses ontologies to deal with the spatial relationship between the moving objects. The ODGIS (Ontology-Driven Geographical Information Systems) framework [6] developed a comprehensive usage of ontologies for gradual classification purposes, focusing on integrating different kinds of geographic information. Kuipers [7] proposed an intelligent cognition approach to space from an artificial intelligence viewpoint, mainly …