Abstract
Understanding and browsing of data or knowledge is still a difficult task especially when users are confronted with large volumes of data. A considerable amount of research has focused on extracting "skyline" points as a restitution tool. Most of the recent work has taken the preferences into account, but existing solutions are not very efficient in terms of information storage, as they store additional information to improve response time to queries. Our proposal, EC2Sky, focuses on two points (1) how to respond effectively to requests like skyline, in the presence of user preferences despite large volumes of data (both in terms of dimensions and preferences) (2) how to restore the most relevant knowledge by underlining the associated trade-offs with the specified preferences.