Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Classement d'objets Skylines dans les bases de donn{é}es

Mickaël Martin-Nevot, Lotfi Lakhal

TL;DR

This work proposes to improve the dp-idp method, inspired by tf-idf, a recent approach computing a score for each Skyline point, by introducing the concept of dominance hierarchy, and introduces the TOPSIS based CoSky method, derived from both information research and multi-criteria analysis.

Abstract

Multi-criteria decision analysis in databases has been actively studied, especially through the Skyline operator. Yet, few approaches offer a relevant comparison of Pareto optimal, or Skyline, points for high cardinality result sets. We propose to improve the dp-idp method, inspired by tf-idf, a recent approach computing a score for each Skyline point, by introducing the concept of dominance hierarchy. As dp-idp does not ensure a distinctive rank, we introduce the TOPSIS based CoSky method, derived from both information research and multi-criteria analysis. CoSky, directly embeddable in DBMS, automatically ponderates normalized attributes using the Gini index, then computes a score using Salton's cosine toward an ideal point. By coupling multilevel Skyline to CoSky, we introduce DeepSky. CoSky and dp-idp implementations are evaluated experimentally.

Classement d'objets Skylines dans les bases de donn{é}es

TL;DR

This work proposes to improve the dp-idp method, inspired by tf-idf, a recent approach computing a score for each Skyline point, by introducing the concept of dominance hierarchy, and introduces the TOPSIS based CoSky method, derived from both information research and multi-criteria analysis.

Abstract

Multi-criteria decision analysis in databases has been actively studied, especially through the Skyline operator. Yet, few approaches offer a relevant comparison of Pareto optimal, or Skyline, points for high cardinality result sets. We propose to improve the dp-idp method, inspired by tf-idf, a recent approach computing a score for each Skyline point, by introducing the concept of dominance hierarchy. As dp-idp does not ensure a distinctive rank, we introduce the TOPSIS based CoSky method, derived from both information research and multi-criteria analysis. CoSky, directly embeddable in DBMS, automatically ponderates normalized attributes using the Gini index, then computes a score using Salton's cosine toward an ideal point. By coupling multilevel Skyline to CoSky, we introduce DeepSky. CoSky and dp-idp implementations are evaluated experimentally.

Paper Structure

This paper contains 36 sections, 16 equations, 8 figures, 11 tables, 7 algorithms.

Table of Contents

  1. Introduction
  2. Cas d'utilisation
  3. Définitions préliminaires
  4. Préférence et dominance
  5. Opérateur Skyline
  6. Classement de Skyline
  7. Méthode dp-idp
  8. Amélioration de dp-idp
  9. Algorithme de la méthode améliorée
  10. Méthode CoSky
  11. I. Normalisation par la somme
  12. II. Pondération avec indice de Gini
  13. III. Détermination du point idéal
  14. IV. Scores avec le cosinus de Salton
  15. V. Classement des résultats
  16. ...and 21 more sections

Figures (8)

  • Figure 1: Les Pokémon du cas d'utilisation
  • Figure 2: Exemple de graphe de hiérarchie de dominance
  • Figure 3: Graphe de hiérarchie de dominance de l'exemple
  • Figure 4: Temps de réponse des différentes solutions
  • Figure 5: Temps de réponse de CoSky SQL (3, 6 et 9 attributs)
  • ...and 3 more figures

Theorems & Definitions (1)

  • definition 1: Hiérarchie de dominance