A preference SQL approach to improve context-adaptive location-based services for outdoor activities

Abstract

Personalized and context-aware information filtering poses a major challenge for the development of meaningful location-based recommendation services, and even more for complex mobile apps for planning outdoor activities. To this end search engine technology has to be extended by sophisticated preference handling capabilities for large amounts of application data, which is typically stored in relational geo-databases like PostGIS. In this paper we propose to employ the Preference SQL query language with its Best-Matches-Only query model as one key enabling technology. Moreover, we introduce a novel location-based preference constructor for querying such geo-databases. We further demonstrate how this extended Preference SQL system can be integrated with an existing commercial web platform for planning outdoor activities such as hiking or mountain bike tours. By performing personalized and context-aware query composition, employing a preference repository attached with social networks and using online information like weather conditions, best matching tour recommendations can be computed with one single query and delivered to mobile smart phones. We are highly confident that forthcoming practical use case evaluations will show the effectiveness and efficiency of our approach for a wide range of application scenarios.