CAdEt – Context-aware Advertising Environment

CAdEt stands for Context-aware Advertising Environment and is the result of a case study which was run with students at the University of Duisburg-Essen, Germany. The idea behind the project is to show advertisements (“ads”) on public displays based on the interests of passersby. Instead of addressing all passersby with the same ads, groups or individuals can be addressed directly. This improves the efficiency of advertisements as potential customers can be notified of new products. Information passersby are interested in can be provided. Persons who are not interested in an ad will not even give it a glance.

This can be done with different technologies. Users are identified with the help of Bluetooth. As each Bluetooth device contains a unique hardware address, it creates the opportunity to inform about the user’s profile. The hardware address can be used to access the individual profiles stored on a server.

Displays have to scan for such addresses and to request profiles from a server. As soon as a group stands in front of a display ads need to be identified which each user is interested in. The system is described in [1].

The initial version was completely based on Java technology and J2ME phones. With the help of J2ME phones users could interact with displays which resulted in adaptions of their personal profile. Users had to leave Bluetooth on their phones switched on to be identified. However, modern phones switch off Bluetooth automatically after a short interval to reduce security risks. Therefore, the initial identification process was not supported anymore.

As a result, the system was redesigned and reimplemented. The system, the server infrastructure and a literature research are explained in my Master project. The new version focused on the supplement of all kind of mobile phones and different kinds of displays. First, a classification of interaction opportunities was accomplished. Based on the results a new system was created allowing different interaction styles, context-aware advertising and much more. The new platform shall be used as a fundamental layer for future research.

[1] Florian Alt, Moritz Balz, Stefanie Kristes, Alireza Sahami Shirazi, Julian Mennenöh, Albrecht Schmidt, Hendrik Schröder and Michael Goedicke: Adaptive User Profiles in Pervasive Advertising Environments. In Proceedings of the 3rd European Conference on Ambient Intelligence (AmI ’09), Austria, 2009