# Efficient skyline query processing with user-specified conditional preference

## TL;DR

This paper introduces a new method for skyline queries that incorporates user preferences using conditional preference networks, improving decision-making efficiency and personalization.

## Contribution

The novel CP-Skyline method integrates user-specified conditional preferences into skyline queries using CP-Nets, enhancing personalization and query efficiency.

## Key findings

- CP-Skyline effectively prunes candidate datasets using user-defined conditional preferences.
- Extensive experiments show significant improvements in skyline quality and query performance.
- The method provides a practical solution for personalized decision support in multi-attribute scenarios.

## Abstract

In the realm of multi-attribute decision-making, the utilization of skyline queries has gained increasing popularity for assisting users in identifying objects with optimal attribute combinations. With the growing demand for personalization, integrating user’s preferences into skyline queries has emerged as an intriguing and promising research direction. However, the diverse expressions of preferences pose challenges to existing personalized skyline queries. Current methods assume that user preferences are too simplistic and do not represent the interdependencies between attributes. This poses a challenge to the existing skyline methods in effectively managing complex user preferences and dependencies. In this article, we propose an innovative and efficient method for skyline query processing, leveraging conditional preference networks (CP-Nets) to integrate specific user’s conditional preferences into the query process, termed as CP-Skyline. Firstly, we introduce a user-defined conditional preference model based on CP-Nets. By integrating user’s conditional preference information, we prune the candidate dataset, effectively compressing the query space. Secondly, we define a new dominance relation for CP-Skyline computation. Finally, extensive experiments were conducted on both synthetic and real-world datasets to assess the performance and effectiveness of the proposed methods. The experimental results unequivocally demonstrate a significant enhancement in skyline quality, and it presents a practical and potent solution for personalized decision support.