Sequential Query Expansion using Concept Graph Date: 2016/4/18 - - PowerPoint PPT Presentation

sequential query expansion using concept graph
SMART_READER_LITE
LIVE PREVIEW

Sequential Query Expansion using Concept Graph Date: 2016/4/18 - - PowerPoint PPT Presentation

Sequential Query Expansion using Concept Graph Date: 2016/4/18 Author: Said Balabeshin-Kordan, Alexander Kotov Source: CIKM16 Advisor: Jia-Ling Koh Speaker: Chih-Hsuan Tzang 1 Introduction Method Experiment Conclusion 2 Introduction


slide-1
SLIDE 1

Sequential Query Expansion using Concept Graph

Date: 2016/4/18 Author: Said Balabeshin-Kordan, Alexander Kotov Source: CIKM’16 Advisor: Jia-Ling Koh Speaker: Chih-Hsuan Tzang

1

slide-2
SLIDE 2

Introduction Method Experiment Conclusion

2

slide-3
SLIDE 3

Introduction Method Experiment Conclusion

3

slide-4
SLIDE 4

Introduction

Goal:

  • ConceptNet

4

slide-5
SLIDE 5

Introduction

Goal:

5

Q: poach preserve wildlife

slide-6
SLIDE 6

Introduction

6

QueryExpansion(LCE) ConceptGraphs SequentialConceptExpansion

twostage- I. InitialSortingofConcepts

  • II. SequentialSelectionofConcepts

Query related concepts

slide-7
SLIDE 7

Introduction Method Experiment Conclusion

7

slide-8
SLIDE 8

Method

8

QueryExpansion(LCE) ConceptGraphs SequentialConceptExpansion

twostage- I. InitialSortingofConcepts

  • II. SequentialSelectionofConcepts

Query related concepts

slide-9
SLIDE 9

Method

9

Latent Concept Expansion (LCE):

  • LCE was designed to incorporate the query

expansion terms from the top retrieved documents into Markov Random Fields-based retrieval models.

slide-10
SLIDE 10

Method

10

Latent Concept Expansion (LCE):

slide-11
SLIDE 11

Method

11

Latent Concept Expansion (LCE):

  • Scoring the document D with respect to query Q
slide-12
SLIDE 12

Method

12

Latent Concept Expansion (LCE):

Q: poach preserve wildlife {“poach”, “preserve”, “wildlife”}

{“poach preserve”, “preserve wildlife”} {“poach preserve”, “preserve poach”, “preserve wildlife”“wildlife preserve”}

slide-13
SLIDE 13

Method

13

QueryExpansion(LCE) ConceptGraphs SequentialConceptExpansion

twostage- I. InitialSortingofConcepts

  • II. SequentialSelectionofConcepts

Query related concepts

slide-14
SLIDE 14

Method

14

Concept Graphs:

  • two way to construct concept Graphs
  • i. use a manually created semantic network, such as
  • ConceptNet. Only considered English concepts.
slide-15
SLIDE 15

Method

15

Concept Graphs:

  • two way to construct concept Graphs
  • ii. use a collection itself. Only unigram concepts are used

in the concept graph in this case. use Hyper-space Analogue to Language(HAL) similarity measure

slide-16
SLIDE 16

Method

16

QueryExpansion(LCE) ConceptGraphs SequentialConceptExpansion

twostage- I. InitialSortingofConcepts

  • II. SequentialSelectionofConcepts

Query related concepts

slide-17
SLIDE 17

Method

17

Sequential Concept Expansion:

  • Step 1: Initial Sorting Concept
slide-18
SLIDE 18

Method

18

Sequential Concept Expansion:

Discard Discard Discard

  • Step 2: Sequential

Selection of Concepts

slide-19
SLIDE 19

19

Sequential Concept Expansion:

  • Step 2: Sequential

Selection of Concepts

slide-20
SLIDE 20

Method

20

Sequential Concept Expansion:

  • Step 2: Sequential

Selection of Concepts

slide-21
SLIDE 21

Introduction Method Experiment Conclusion

21

slide-22
SLIDE 22

Experiment

22

statistics of the collections:

slide-23
SLIDE 23

Experiment

23

slide-24
SLIDE 24

Experiment

24

Remove features MAP:

slide-25
SLIDE 25

Experiment

25

slide-26
SLIDE 26

Experiment

26

slide-27
SLIDE 27

Experiment

27

slide-28
SLIDE 28

Experiment

28

QL- Query Likelihood retrieval model with Dirichlet prior smoothing RM- Relevance Model SDM- Sequential Dependence Model LCE- d Latent. Concept Expansion pseudo-relevance feedback (PRF)

slide-29
SLIDE 29

Introduction Method Experiment Conclusion

29

slide-30
SLIDE 30

Conclusion

30

  • The main contribution of this work:
  • A two-stage method for sequential selection of efgective

concepts for query expansion from the concept graph.

  • The optimization problem of the proposed method:
  • Objective: having least possible number of candidate

concepts.

  • Constraint: achieve a given precision of retrieval results.
  • Stages of the proposed method:
  • stage 1: sort the candidate concepts
  • stage 2: sequentially select expansion concepts