Estimation of the maximum domination value in multi-dimensional data sets

Abstract

The last years there is an increasing interest for query processing techniques that take into consideration the dominance relationship between objects to select the most promising ones, based on user preferences. Skyline and top-k dominating queries are examples of such techniques. A skyline query computes the objects that are not dominated, whereas a top-k dominating query returns the k objects with the highest domination score. To enable query optimization, it is important to estimate the expected number of skyline objects as well as the maximum domination value of an object. In this paper, we provide an estimation for the maximum domination value for data sets with statistical independence between their attributes. We provide three different methodologies for estimating and calculating the maximum domination value, and we test their performance and accuracy. Among the proposed estimation methods, our method Estimation with Roots outperforms all others and returns the most accurate results. © 2010 Springer-Verlag.