SIFT - PowerPoint PPT Presentation

1 / 24
About This Presentation
Title:

SIFT

Description:

????? ????? SIFT. ???? ???? ?????. ????? ?????. ????? ?????? ????? ??? ??????? ??????? ... ????? ???? ?? SIFT (?????) ???? ???? ???? ??????? ?? ????? G(m,n) ... – PowerPoint PPT presentation

Number of Views:39
Avg rating:3.0/5.0
Slides: 25
Provided by: ceSh
Category:
Tags: sift | sift

less

Transcript and Presenter's Notes

Title: SIFT


1
????? ??? ?????? ???????
  • ????? ????? SIFT
  • ???? ???? ?????

2
????? ?????
  • ????? ?????? ????? ??? ??????? ???????
  • ????? ??? ?????? ???????
  • ????? ????? SIFT
  • ?????? ???? ????? ?? SIFT
  • ????? ???? SIFT
  • ??????? SIFT

3
????? ?????? ????? ??? ??????? ???????
  • ???????? ???? ?? ????? IR
  • ????????? ????? ?????
  • ????????? ????? ??? ? ?? ??
  • ?????? ??? ?? ? ??? ? ????
  • ?????? ?? ???? ????? (?????? NLP)
  • ?? ??????? ????
  • ??????????
  • ???????

4
?????? ?? ???? ?????? ?????? ???????
5
IR ?? ?????? ?? IF
Grand Challenge
Filtering
Information Source change rate
Retrieval
information need change rate
6
????????? ???? ?? ?? ????? ?????? ???????
  • ??? ???? ???????
  • ????
  • ???????
  • ??????
  • ?????
  • ??????
  • ??? ??????

??? ????
??????
?????
7
?????? ??? ?? ????? ??????
Human Judgment
j
User Interest space I
Document space D
Info. Need
Document
p
d
Representation space R
profile
Representation
c
Comparison Function
n0,1
8
?? ?????? ???? ?? ??????
  • ?????-???? (content-based)
  • SIFT
  • InfoScope
  • ??????? (Social)
  • Tapestry
  • GroupLens

9
????? ??? ???? ?? ?????
  • ??? ???? ?????
  • ???? (SIFT)
  • ????
  • ??? ???? ??? ? ?? ??
  • Boolean
  • Vector Based
  • ???? ?????

10
?????? ??? SIFT
11
???????? ???? ????? SIFT
  • ????? ???? ??? ????? ? profile ??
  • ??????? ?? ????? ???????
  • ????? ?????? ????? ?? ??? ????? ??? ?????

12
????? ??? ?????
  • Mailing list
  • Boston Community Information System
  • ?????? ????
  • Tapestry
  • Social
  • Relational Query

13
???? ???? ?? ????? SIFT
  • ?? ????? ?? ?? ??? profile ????
  • Profile
  • ??? ??? ? ??
  • ?????? ????? ???? ?? ?????
  • ??? profile
  • ????? ?????????? ?????

14
???? ???? ?? ????? SIFT (?????)
  • ?????? ??? ? ?? ??
  • Boolean
  • Duamish not Microsoft
  • VSM
  • lt0.3 0 0.5 0.6 0 gt
  • ltFlower,0.3 Red,0.5gt
  • Relevance feedback

15
???? ???? ?? SIFT
  • ??? ??? ??
  • ?????? ????? ??? ??? ? ??? ? ??
  • ??????? ?? ?? Relevance Threshold

16
?????? ????? ???? ????
  • Brute Force (BF) Method
  • ?????? ??? ????? ?? ???? ??? ? ?? ?? ????? ???
  • Query Indexing (QI) Method
  • Inverted List
  • 2 ????? ?? ????? ??? ???? ?? ???
  • Threshold
  • Scores

17
?????? ????? ???? ???? (?????)
18
?????? ????? ???? ???? (?????)
  • Selective Query Indexing (SQI)
  • ??? ? ?? ?? ?? ?????? ???? ?? ??? ?? ???? ?????
    ?? ????
  • ???? ???
  • ????? insignificant sub-vector
  • ???? ISV ?? ?? ?????? ?????
  • Low idf
  • Most insignificant
  • ????? ??? sv ?? ?? ????? ????
  • Sim(d,q)lt q
  • ????? ?? ?? ???? idf ???? ?? ????

19
?????? ????? ???? ???? (?????)
20
???? ???? ??? ??????? ?????
  • ?????? ????? ????? ? ??? ? ?? ??? ?????
  • Vocabulary
  • Stop words
  • Stemming
  • Threshold
  • SQI?? Threshold ??? ???? ??? ???? ?? ???
  • ???? SIFT ?? QI ??????? ??

21
????? ???? ?? SIFT
  • ?????? ????? ????
  • ?????? ??? ????? ? ??? ? ????
  • ???? ?? ?? ????? ?? ?????? ??????? ????
  • ????? quorum ,
  • Document quorum D
  • Profile quorum P
  • pdn1
  • M(d,p)
  • ????? ?????? ???? ?? ????? ???? ?? ????? ???? ????

22
????? ???? ?? SIFT (?????)
  • ???? ???? ???? ??????? ?? ????? G(m,n)
  • ??????? ?? ?????? ??? ????? ???? ?? d ???? p
    ????
  • ??? ??????-????? ?????
  • MM ,MG,GM,GG
  • ?? SIFT ?? G ??????? ??? ??? G(1,4),G(2,2),G(4,1)

23
??? ???? ???? ?? ????? ??? ??????
  • ?? ???? ?? ????? ???????? ?? ???? ????? ???????
    ?? ?????? ???? ??? ??????? ??????? ???? ?? ????
  • ??????? ?? ?????? ??? ???? ?? Profile ??
  • ??? ????
  • ?? ???? ???????
  • ??????? ?? ??????? Profile
  • ????? ?????? ? ??????? ?????? ???

24
?????
  • A Conceptual Framework for Text Filtering
  • Douglas W. Oard Gary Marchionini
  • Information filtering and Information retrieval
    two sides of the same coin.
  • Nicholas J. Belkin W. Bruce Croft
  • The SIFT Information Dissemination Systems
  • Tak W. Yan, Hector Garcia Molina
Write a Comment
User Comments (0)
About PowerShow.com