Multi-objective optimization: basic approaches and moving beyond them through flexible skyline queries

TL;DR

This paper reviews multi-objective optimization in databases, compares basic approaches, and introduces flexible skylines as an innovative framework that overcomes limitations of traditional methods.

Contribution

It presents the flexible skylines framework, offering a superior and more adaptable solution for multi-objective optimization in database systems.

Findings

Flexible skylines outperform basic approaches.

Flexible skylines overcome main drawbacks of traditional methods.

The framework enhances data filtering and extraction capabilities.

Abstract

The area of scientific research that deals with the simultaneous optimization of several (possibly conflicting) criteria is named multi-objective optimization. The ability to efficiently filter and extract interesting data out of large datasets is one of the key tasks in modern database systems. This paper provides a general overview of the most common approaches employed to handle the problem in the field of databases, and describes a novel framework named flexible skylines. After analyzing the main differences between single and multi-optimization problems, I will discuss the three main basic approaches used to handle multi-optimization problems: lexicographic approach, top-k queries and skylines. Each methodology will be discussed, analyzing the pros, the range of applicability and the main issues, which motivate the need to introduce the flexible skylines innovative framework. A review of this approach will show its superiority with respect to the basic approaches, as well as the capability to overcome the majority of their drawbacks.