Personalised recommendation addresses the need for a dynamic, contextualised delivery of results based on the qualifications and preferences of a user. Clearly, delivering high-quality personalized recommendations depends on the skills of a user, her current context and prerequisites for understanding the delivered content. For instance, a teacher preparing a lesson about the Napoleon wars will get multi-media contents and documents depicting certain events and historical persons relevant to the Napoleon wars. She will not be recommended resources that are in a language she or her pupils do not know.
Establishing such high-quality recommendations is a big challenge. In order to achieve this objective, memory organisations must be given the ability to add their expertise and domain knowledge in the recommendation process and to advance their knowledge on user preferences and user needs, while at the same time protecting users from any privacy infringements.
Since large-scale recommendation will become cost-intensive, it is unlikely that one institution will be able to run a single recommender. Therefore we aim to create a recommender network, in which every recommender is specialised on particular content and a subset of a user group. Aggregating recommendation results over heterogeneous sources will become challenging in terms of accuracy and timeliness.
EEXCESS will research and develop personalised, contextualised and decentralised recommendation technology capable to scale to large numbers of users. Personalisation and contextualisation target the satisfaction of users, whereas decentralisation focuses on the potential uptake of the EEXCESS framework by different organisations while retaining the capabilities of a centralised recommender.
|The key objectives of EEXCESS:|
|1. Enrichment of content|
|2. Personalised recommendation|
|3. Privacy Preservation|