A Chinese Wall Approach for Anonymous Recommendation in a Multi-Dimensional-Personalisation Scenario
Multi-Dimensional-Personalisation, especially if it spans between the online and offline world (i.e., in aSchilke SW, Bleimann U, Furnell SM, Phippen AD
mobile environment), requires a pro active recommendation service which is not only based on the interest of the users
but also on the location of the user and a temporal component describing the "when". As users are very privacy
conscious such a service has to take care of providing privacy while delivering services. In order to achieve this a "Chinese Wall" approach is proposed based on a middleman to allow push services based recommendation without
sacrificing privacy. By doing so recommernders can select their target audience via an anonymous profile and still reach
their target audience without gaining access to any personal data of the user.