ABSTRACT

Skyline queries have recently attracted a lot of attention for its intuitive query formulation. It can act as a filter to discard sub-optimal objects. However, a major drawback of skyline is that, in datasets with many dimensions, the number of skyline objects becomes large and no longer offer any interesting insights. To solve the problem, k-dominant skyline queries have been introduced, which can reduce the number of skyline objects by relaxing the definition of the dominance. However, sometimes, a kdominant skyline query may retrieve too few objects to analyze. This paper addresses the problem of k-dominant skyline for high dimensional dataset. In addition, we extend the notion of k-domination by defining extended k-dominant skyline, which retrieves neither too many nor too few objects. We propose algorithms for k-dominant and extended kdominant skyline computation. An extensive performance evaluation using both real and synthetic datasets demonstrated that our proposed methods are efficient and scalable.