20
SCALABLE SKYLINE COMPUTATION USING OBJECT-BASED SPACE PARTITIONING Shiming Zhang Nikos Mamoulis David W. Cheung sigmod 2009 1

Scalable Skyline Computation Using Object-based Space Partitioning

  • Upload
    vila

  • View
    34

  • Download
    0

Embed Size (px)

DESCRIPTION

Scalable Skyline Computation Using Object-based Space Partitioning. Shiming Zhang Nikos Mamoulis David W. Cheung sigmod 2009. Outline. Introduction Object-based Space Partitioning Recursive Object-based Space Partitioning Left-Child/Right-Sibling Skyline Tree OSPSOnSortingFirst - PowerPoint PPT Presentation

Citation preview

Page 1: Scalable Skyline Computation Using Object-based  Space Partitioning

1

SCALABLE SKYLINE COMPUTATION USING OBJECT-BASED SPACE PARTITIONING

Shiming Zhang

Nikos Mamoulis

David W. Cheung

sigmod 2009

Page 2: Scalable Skyline Computation Using Object-based  Space Partitioning

2

OUTLINE

Introduction Object-based Space Partitioning Recursive Object-based Space Partitioning Left-Child/Right-Sibling Skyline Tree

OSPSOnSortingFirst OSPSOnPartitioningFirst

FilterDominatedPartitions

Experimental Results Conclusions

Page 3: Scalable Skyline Computation Using Object-based  Space Partitioning

3

INTRODUCTION(1)

Skyline queries are useful in multi-criteria decision making applications that involve high dimensional and large datasets.

There is a number of methods that operate on pre-computed indexes on the data.

Compare each accessed point with the skyline points found so far.

Page 4: Scalable Skyline Computation Using Object-based  Space Partitioning

4

INTRODUCTION(2)

Object-based space partitioning(OSP) scheme, which recursively divides the d-dimensional space into separate partitions w.r.t. a reference skyline object.

Organizes the current skyline points in a search tree.

Object o dominates another object o' iff o is not worse than o' in all dimensions and better than o' in at least one dimension.

Page 5: Scalable Skyline Computation Using Object-based  Space Partitioning

5

NOTATION

Page 6: Scalable Skyline Computation Using Object-based  Space Partitioning

6

OBJECT-BASED SPACE PARTITIONING

reference skyline

Page 7: Scalable Skyline Computation Using Object-based  Space Partitioning

7

RECURSIVE OBJECT-BASED SPACE PARTITIONING

reference skyline

Page 8: Scalable Skyline Computation Using Object-based  Space Partitioning

8

WHY CAN SAFELY SKIP?

Skip all incomparable partitions according to Corollary 1

Page 9: Scalable Skyline Computation Using Object-based  Space Partitioning

9

LEFT-CHILD/RIGHT-SIBLING SKYLINE TREE

Page 10: Scalable Skyline Computation Using Object-based  Space Partitioning

10

LEFT-CHILD/RIGHT-SIBLING SKYLINE TREE

Page 11: Scalable Skyline Computation Using Object-based  Space Partitioning

11

LCRS TREE GROWTH

Page 12: Scalable Skyline Computation Using Object-based  Space Partitioning

12

TRACE

Page 13: Scalable Skyline Computation Using Object-based  Space Partitioning

13

OSP SKYLINE ALGORITHMS 1

Page 14: Scalable Skyline Computation Using Object-based  Space Partitioning

14

OSP SKYLINE ALGORITHMS 2

Page 15: Scalable Skyline Computation Using Object-based  Space Partitioning

15

OSP SKYLINE ALGORITHMS

Page 16: Scalable Skyline Computation Using Object-based  Space Partitioning

16

OSP SKYLINE ALGORITHMS

Page 17: Scalable Skyline Computation Using Object-based  Space Partitioning

17

EXPERIMENTAL EVALUATION

Three types of synthetic datasets anti-correlated (AC)

NBA uniform and independent (UI)

Household correlated (CO)

Color

Page 18: Scalable Skyline Computation Using Object-based  Space Partitioning

18

EXPERIMENTAL RESULTS

Page 19: Scalable Skyline Computation Using Object-based  Space Partitioning

19

EXPERIMENTAL RESULTS

Page 20: Scalable Skyline Computation Using Object-based  Space Partitioning

20

CONCLUSIONS

Proposed an efficient set of skyline evaluation algorithms that are based on the idea of organizing the discovered skyline points in a tree.

Each candidate skyline object only needs to be compared for dominance with a small subset of the existing skyline points. (skip incomparable sets )

Makes our solutions scalable to the dimensionality, a feature that all previously proposed skyline algorithms lack.