atish das sarma, ashwin lall, danupon nanongkai, jun xu 1 georgia tech vldb 2009

74
Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Randomized Multi- pass Streaming Skyline Algorithm Georgia Tech VLDB 2009

Upload: preston-kelly

Post on 31-Mar-2015

217 views

Category:

Documents


4 download

TRANSCRIPT

Page 1: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

1

Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu

Randomized Multi -pass Streaming Skyline Algorithm

Georgia Tech

VLDB 2009

Page 2: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

2

In one sentence ….

Page 3: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

3

“We develop a streaming algorithm

Page 4: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

4

“We develop a streaming algorithm for skyline problem

Page 5: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

5

“We develop a streaming algorithm for skyline problem with near-optimal worst-case guarantee.”

Page 6: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

6

What is skyline?

Page 7: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

7

Hotel Price DistanceAthena $97 2.9 km

Park & Suites $124 3.6 km

Hotel du Helder $76 3.8 km

de la Cité Concorde $220 0.67 km

Mercure Carlton Lyon $163 3.0 km

I want a cheap hotel

nearby

Page 8: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

8

Hotel Price DistanceAthena $97 2.9 km

Park & Suites $124 3.6 km

Hotel du Helder $76 3.8 km

de la Cité Concorde $220 0.67 km

Mercure Carlton Lyon $163 3.0 km

I want a cheap hotel

nearbydo

min

ates

Page 9: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

9

Hotel Price DistanceAthena $97 2.9 km

Park & Suites $124 3.6 km

Hotel du Helder $76 3.8 km

de la Cité Concorde $220 0.67 km

Mercure Carlton Lyon $163 2.9 km

I want a cheap hotel

nearbydo

min

ates

Page 10: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

10

Price

Distance

de la Cite

Park & Suites

du HelderAthena

Mercure

Page 11: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

11

Price

Distance

de la Cite

Park & Suites

du HelderAthena

Mercure

Page 12: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

12

Problem definition

• Given distinct d-dimensional points• (a1, …, ad) dominates (b1, …, bd) if ai ≤ bi for all i

and ai’ < bi’ for some i’• Skyline = set of undominated points

dominatesSkyline = { (1, 3) , (3, 2) }

(5,2)

(1,3)

(3,2)

Example(1, 3) , (5, 2) , (3, 2)

Page 13: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

13

Skyline algorithms

RAM Disk (External)

Preprocessing Non-preprocessingBBS Papadias et al. SIGMOD’03NN Kossman et al. VLDB’02

DD&C Kung et al. FOCS’ 75LD&C Bently et al. JACM’78, FLET Bently et al. SODA’90,

SD&C Borzsonyi et al. ICDE’01,BNL Borzsonyi et al. ICDE’01, SFS Chomicki et al. ICDE’03, LESS Godfrey et al. VLDB’05

Page 14: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

14

Our Goal“Non-preprocessing external

algorithm with worst-case guarantee”

What is the model of external algorithms?

Page 15: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

15

CPU process ≠ I/OSequental I/O ≠ Random I/O

Models for external algorithms

Multi-pass Streaming

Model

# of random I/O’s = # of passes

Streaming model naturally forces us to minimize the number of random I/O’s

Page 16: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

16

What is multi-pass stream?

Page 17: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

17

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

Small RAM

Huge Harddisk

Page 18: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

18

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

Small RAM

Huge Harddisk

Page 19: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

19

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

Small RAM

Huge Harddisk

Page 20: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

20

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

Small RAM

Huge Harddisk

Page 21: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

21

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

Small RAM

Huge Harddisk

2nd pass

Page 22: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

22

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

Small RAM

Huge Harddisk

3rd pass

Page 23: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

23

Our Goal

“Non-preprocessing external algorithm with worst-case guarantee”

streaming

Page 24: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

24

Main resultsTheory

RAND: Almost optimal multi-pass streaming algorithm for skyline

O(log n) passes & O(m) space

n = # of points and m = skyline size

1 pass needs Ω(n) space

• RAND uses O(log n) passes & O(m) space• Every algorithm that uses 1 pass needs Ω(n) space

Next: RAND algorithmLater: Experimental result

Page 25: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

25

RAND algorithm

Page 26: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

26

Algorithms: Main Idea

Suppose m is known.Theorem: In 3 passes and m space, we

can find skyline points that “dominate” at least n/2 points, with high probability

Page 27: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

Eliminate-Points algorithm

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)27

Page 28: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

28

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 29: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

29

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 30: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

30

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)(3, 4)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 31: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

31

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 32: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

32

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 33: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

33

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 34: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

34

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 35: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

35

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 36: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

36

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 37: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

37

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 38: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

38

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 39: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

39

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 40: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

40

Eliminate-Points algorithm

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)(3, 3)

1. Sample x=2m ln(mn log n) points p1, p2, …, px

2. Go through the stream,Replace each pi by a point dominating it

3. For each pi, delete pi and all points it dominates

Output p1, p2, …, px and repeat

Page 41: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

41

Analysis

Theorem: Eliminate-Points algorithm deletes at least n/2 points with high probability

Page 42: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

42

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

Page 43: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

43

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

Page 44: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

44

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

Note: There will be m trees, each rooted by a skyline point

Page 45: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

45

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

Page 46: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

46

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

(3, 3)

Page 47: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

47

4, 4

Analysis

• Claim: The tree that some element is sampled will be deleted

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

1, 5 3, 3

3, 4 4, 34, 5

(3, 3)

Page 48: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

48

Analysis

• There are m trees, each rooted by a skyline point

1 2 mm-1

Page 49: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

49

Analysis

• There are m trees, each rooted by a skyline point

1 2 mm-1

Page 50: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

50

Analysis

• Big tree has bigger chance of being sampled… and deleted

1 2 mm-1

Page 51: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

51

Analysis

• If enough points are sampled, every tree that is “big enough” will be deleted

1 2 mm-1

Page 52: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

52

Analysis

Lemma: With high probability, all trees of size n/(2m) are deleted

• We delete n/2 points in total1 2 mm-1

Page 53: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

53

Extending to RAND• Recall: If we know m then we can delete n/2 points

in 3 passes• If m is known, we can find skyline in O(log n)

passes with high probability– We delete n/2 points every 3 passes

• m is not known– Guess m by “doubling trick” – Additional O(log m) passes

• Fixed-window case – Memory space is limited

• Random I/O’s, Sequential I/O’s and Number of comparisons have to be analyzed separately

Page 54: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

54

Main resultsTheory

RAND: Almost optimal multi-pass streaming algorithm for skyline

O(log n) passes & O(m) space

n = # of points and m = skyline size

1 pass needs Ω(n) space

• RAND uses O(log n) passes & O(m) space• Every algorithm that uses 1 pass needs Ω(n) space

Page 55: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

55

TheoryRAND: Almost optimal multi-pass streaming algorithm for skyline

O(log n) passes & O(m) space

n = # of points and m = skyline size

1 pass needs Ω(n) space

Algorithms comparison w = window (memory) size

Main results

Algorithm Random I/O’s Sequential I/O’s ComparisonsBNL(w) Q(min{w, n/w}) Q(min{w, n2/w}) Q(dmin{wmn, n2})LESS(w) Q(n logw (n/w)) Q(mn/w) Q(dmn+n log n)

RAND(w) O(m log (n/w)) O(mn/w) O(dmn)

Page 56: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

56

Main resultsExperiment RAND BNL & LESSvs

Average case

Worst case

We try several datasets in the literature …

Correlate, Anti-correlated, Independent,Island, House, NBA, Color

Page 57: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

57

Average case- No clear winner between BNL and LESS- RAND is always close to the winner

Experimental Results

RAND BNL & LESS

Page 58: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

Experimental Results

58RAND

“Worse”: After sorting by decreasing first coordinate- RAND is the most robust and usually fastest

BNL & LESS

Page 59: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

Experimental Results

59RAND BNL & LESS

“Even Worse”: After sorting by “entropy”

Page 60: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

Summary

60

(1, 2) (3, 7) (5, 3) (2, 5) (4, 1) (9, 9)

60

RAND BNL & LESS

Average case

Worst case

Disk Stream

1 2 mm-1Random Sampling RAND

Experiment

Page 61: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

61

Extensions• Distributed skyline algorithm• Derandomize the algorithm for 2D case• Skyline for partially ordered sets (posets)Open problems• Develop algorithm on Parallel Disk Model

(PDM) and Cache Oblivious model• Extend the techniques to pre-processing

algorithm• Is O(log n) passes the best possible?

Summary

Page 62: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

62

Thank you

Page 63: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

63

Appendix

Page 64: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

64

Charts for average case

Page 65: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

65

Page 66: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

66

The lower bound

Theorem: Any randomized one-pass algorithm with space at most n/2 succeeds with probability at most 1/2

Proof- Random unique survivor- 2 points come at the end- If space <= n/2 then will fail if didn’t store survivor in the memory

Page 67: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

67

Proof of Claim

Page 68: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

68

Proof of Claim

• Claim: The tree that some element is sampled will be deleted

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

(3, 3)

Page 69: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

69

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

Page 70: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

70

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(4, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

(3, 4)

3, 4

Page 71: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

71

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

(3, 3)

3, 4

3, 3

Page 72: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

72

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

(3, 3)

3, 4

3, 3

Page 73: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

73

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

(3, 3)

3, 4

3, 3

Page 74: Atish Das Sarma, Ashwin Lall, Danupon Nanongkai, Jun Xu 1 Georgia Tech VLDB 2009

74

Analysis

• Draw trees: Each point points to its first dominating point

(1, 5), (3, 4), (4, 5), (4, 3), (3, 3), (4,4)

(3, 4)

1, 5 3, 3

3, 4 4, 3

4, 4

4, 5

4, 4

(3, 3)

3, 4

3, 3