Flexible skylines, regret minimization and skyline ranking: a comparison to know how to select the right approach

Abstract

Recent studies pointed out some limitations about classic top-k queries and skyline queries. Ranking queries impose the user to provide a specific scoring function, which can lead to the exclusion of interesting results because of the inaccurate estimation of the assigned weights. The skyline approach makes it difficult to always retrieve an accurate result, in particular when the user has to deal with a dataset whose tuples are defined by semantically different attributes. Therefore, to improve the quality of the final solutions, new techniques have been developed and proposed: here we will discuss about the flexible skyline, regret minimization and skyline ranking approaches. We present a comparison between the three different operators, recalling their way of behaving and defining a guideline for the readers so that it is easier for them to decide which one, among these three, is the best technique to apply to solve their problem.