Preference Analytics in EXASolution

Abstract

Skyline queries and the more general concept of preferences are wellknown in the database community and there are many academic approaches for the computation of the best-matching objects. Furthermore, data analytics and multicriteria optimization play an important role in Business Intelligence where it facilitates optimal decision making. Preference Analytics is the combination of preferences and data analytics. SKYLINE is EXASOL's implementation of Preference Analytics in its commercial database management system EXASOLUTION. In this paper, we present SKYLINE from a user's perspective, describe algorithmic design decisions, and discuss its implementation in a distributed and parallel environment. The paper closes with an empirical evaluation of the system based on a number of preference queries over the TPC-H dataset using different scale factors.