Skyline queries are well-known for their intuitive query formalization and easy to understand semantics for selecting the most interesting data objects from large data sets. They naturally fill the gap between set-based queries using strict predicates and only few personalization options and rank-aware database retrieval, offering a high degree of personalization at the cost of very complex query formalization. Thus, skyline queries enjoyed great popularity in the database personalization research community. Unfortunately, the simplicity and elegance of the query paradigm come at high costs: skyline queries often suffer from a problem usually known as “curse of dimensionality”. With the increasing number of query attributes, the size of skyline result sets grows exponentially and the results are thus seldom useful or manageable by users –result sets containing 30%-50% are commonly heard of. This problem severely hinders the practical application of the skyline paradigm.
During the course of this thesis, the concept of trade-off skylines has been incrementally developed and successfully published on numerous international conferences and journals. Trade-off skylines approach the curse of dimensionality by eliciting additional user information in form of intuitive trade-offs. This additional information can be used to compensate between certain characteristics of database objects in order to focus the skyline result sets. Ultimately, this will lead to more manageable and useful query results, and thus alleviating one of the most severe problems of the Skyline paradigm.
In this cumulative doctoral thesis, the problem of unmanageable large Skyline query result sets is addressed and a solution based on cooperative user trade-offs is developed. In the following, after a short introduction to the area of Skyline queries, the relevant papers published during the course of this thesis are summarized and discussed.