Inhalt des Dokuments
Es gibt keine deutsche Übersetzung dieser Webseite.
Master Thesis: Supporting Geo-Recommendations in Location-based Community Services based on Background Tracking Information
Supporting Geo-Recommendations in Location-based Community Services based on Background Tracking Information
With the increasing number of location-based services, context-aware recommender systems become more and more relevant when recommending content items, such as products, restaurants or shops. Contextual data (e.g., location, time of day or weather) is a promising information source to exploit in order to generate more precise recommendations that not only fit to a user’s profile and ratings given to those content items by a community, but also on the contextual situation the user is in. However, not all kinds of content items are suitable for context-based recommendations. Moreover, the automatic detection of some context parameters (e.g., the mood of a user or companions) turns out to be very difficult and therefore the scenarios, in which those context data is used, seem not very applicable in real business services.
The location information, on the other hand, is the most important, trustworthy and automatically detectable context that can be utilized in order to create location-based recommendations: Users get locations recommended, not only based on their interests and ratings, but also on context information that can be directly derived by the location information like dwell time, entry time, weather or user activity. Recommendations that can be generated by this approach include the following examples:
- People who frequently visit the restaurant A are also frequent visitors of restaurants B and C.
- Many people who visit the bar A on a Saturday evening join the discotheque B afterwards.
- People on average spend 10 minutes less time at the haircutter A than at other haircutters.
The main objective of this thesis is the design and implementation of a new location-based recommendation algorithm that incorporates location information and other depending contextual information into the recommendation process in order to create recommendations for locations. The feasibility of the concepts developed within this thesis has to be demonstrated by a prototype.
- Strong practical relevance
- Proficiency in software development
- Ideally first experience with mobile platforms
- Interest to work in the field of Recommender
Systems, Context-aware Computing and Location-based Services
Supervisor: Prof. Dr. Axel Küpper , Abdulbaki Uzun , Ulrich Bareth 
Duration: 6 months
10587 Berlin, Germany
Phone: +49 30 8353 58811
Fax: +49 30 8353 58409