Sei sulla pagina 1di 8

ABSTRACT

Most existing Web search engines return a list of search results based on a users
query but ignore the users specific interests and/or search context. Therefore, the
identical query from different users or in different contexts will generate the same
set of results displayed in the same way for all users, a so called one-size-to-all
approach. Furthermore, the number of search results returned by a search engine
is often so large that the results must be partitioned into multiple result pages. In
addition, individual differences in information needs, polysemy i.e., multiple
meanings of the same word, and synonymy pose problems in that a user may
have to go through many irrelevant results or try several queries before finding the
desired information. Therefore personalization of search results has become a
necessity.
Over the last decade, the World Wide Web and Web search engines
have fundamentally transformed the way people share information. Recently, a
new form of publishing and locating information, known as online social
networking, has become very popular. While numerous studies have focused on
the hyperlinked structure of the Web and have exploited it for searching content,
few studies, if any, have examined the information exchange in online social
networks. This paper presents a method which exploits the user data in the social
networking site and constructs a user profile based on it. The user profile is
constructed using the likes done by user. This user profile is used to obtain
personalized result for the given search query.

CONTENTS
ABSTRACT

LIST OF TABLES

LIST OF FIGURES

vi

LIST OF ABBREVIATIONS

CHAPTER

TITLE

viii

PAGE NO.

NO.
1

INTRODUCTION

1.1

OVERVIEW

1.2

LITERATURE REVIEW

1.2.1 Web Search Engine

1.2.2 Social Networking

1.2.3 Personalization

1.2.4 Document Clustering

1.2.5 Personalized Search Engine

LIMITATIONS AND MOTIVATIONS

1.3.1 Limitations

1.3.2 Motivations

ORGANISATION OF THE PROJECT WORK

1.3

1.4
2

AIM AND SCOPE OF PRESENT INVESTIGATION

2.1

SCOPE OF THE PROJECT

2.2

OBJECTIVES

2.3

EXISTING SYSTEMS

2.4

PROBLEM STATEMENT

10

2.5

OVERVIEW OF PROPOSED SYSTEM

10

EXPERIMENTAL METHODS AND ALGORITHM USED

12

3.1

HARDWARE SPECIFICATIONS

12

3.2

SOFTWARE SPECIFICATIONS

12

3.2.1 Windows 7

12
2

3.2.2 Chrome

12

3.2.3 JAVA

13

3.2.4 JSP

13

3.2.5 PHP

14

3.2.6 MySQL

14

3.2.7 Tomcat 8

14

3.2.8 Carrot2 API

15

3.3

PROPOSED SYSTEM

15

3.4

EXPERIMENTAL SETUP

16

3.5

SYSTEM DESIGN

17

3.5.1 Modular Design

18

3.5.1.1 Search Result Retrieval

18

3.5.1.2 Clustering Search Results

20

3.5.1.3 Facebook Data Retrieval

21

3.5.1.4 Category Classification

21

3.5.1.5 User Profile Construction

23

3.6 PROPOSED ALGORITHM

24

3.6.1 Flow Chart

25

3.6.2 Pre-processing

26

3.6.3 Frequent Phase Extraction

26

3.6.4 Cluster Label Induction

27

3.6.5 Cluster Content Discovery

28

3.6.6 Final Cluster Formation

29

RESULTS, DISCUSSIONS AND PERFORMANCE

30

ANALYSIS
4.1

GENERAL

30

4.2

RESULTS AND DISCUSSIONS

30

4.2.1 User Search

30

4.2.2 Search Results

31

4.2.3 Cluster Label Indexing

31

4.2.4 App Id and Key

32

4.2.5 Categorization on History Of Likes

33

4.2.6 User Likes

33

4.2.7 User Profile Construction

34

4.3

4.2.8 Re-Ranked Results

36

4.2.9 Cluster Formation

37

PERFORMANCE ANALYSIS

38

4.3.1 Test Cases

38

4.3.1.1 Search Results Retrieved

38

4.3.1.2 User Profile1

41

4.3.1.3 User Profile2

45

4.3.2 Bing Vs Personalized Engine

49

SUMMARY AND CONCLUSION

51

5.1

SUMMARY

51

5.2

CONCLUSION

51

REFERENCES

53

APPENDIX SAMPLE CODE

56

PUBLICATION

61

LIST OF TABLES

TABLE NO.
3.1

TABLE NAME
EXAMPLES OF KEY TERM EXTRACTION

PAGE NO.
29

LIST OF FIGURES

FIGURE NO.

FIGURE NAME

PAGE NO.

1.1

Web Search Engine

3.1

Clustering Architecture

16

3.2

User Profile Constructions

17

3.3

Search Engine Architecture

17

3.4

Functional Architecture

18

3.5

Retrieval of Search Results

20

3.6

FB Data Retrieval Steps

22

3.7

User Interest Score Computation

24

3.8

Flow of Lingo Clustering

25

4.1

User Search Query Page

30

4.2

Example Search Results for the


Query Schindler

31

4.3

Example Clusters for the Query Schindler

32

4.4

Facebook app depicting app id and secret code

32

4.5

Like Based User Profile Depicting History of Likes

33

FIGURE NO.

FIGURE NAME

PAGE NO.

4.6

List of Users Likes

34

4.7

User Profile

35

Pie Chart Representing Recent history


4.8

Like Based User Profile


6

35

Pie Chart Representing Recent History


4.9

Like Based User Profile


Re-Ranked Search Results Based On

36

4.10

User Profile
Re-Ranked Search Clusters After Re-

37

4.11

Ranking

37

4.12

Experimental- Like History User Profile 1

44

4.13

Experimental- Recent Like History User Profile 1

44

4.14

User Profile 2 Recent and History of Likes Chart

44

4.15

Bar Chart Depicting Interest Relevance

49

LIST OF ABBREVIATIONS

ABBREVIATION

EXPANSION

TF

TERM FREQUENCY

IDF

INVERSE DOCUMENT
FREQUENCY
7

API

APPLICATION PROGRAMMING
INTERFACE

WDC

WEB DEVELOPMENT COMPANY

DMOZ

DIRECTORY.MOZILLA.ORG

IR

INFORMATION RETRIEVAL

PLSA

PROBABILISTIC LATENT
SEMANTIC ANALYSIS

BSD

BERKELY SOFTWARE
DISTRIBUTION

Potrebbero piacerti anche