WWW.ABSTRACT.XLIBX.INFO
FREE ELECTRONIC LIBRARY - Abstract, dissertation, book
 
<< HOME
CONTACTS



Pages:   || 2 | 3 | 4 | 5 |   ...   | 15 |

«by Jong Do Park B. A., Chung-Ang University, 1996 M. A., Chung-Ang University, 1998 Submitted to the Graduate Faculty of School of Information ...»

-- [ Page 1 ] --

AUTOMATED QUESTION TRIAGE FOR SOCIAL REFERENCE:

A STUDY OF ADOPTING DECISION FACTORS FROM DIGITAL REFERENCE

by

Jong Do Park

B. A., Chung-Ang University, 1996

M. A., Chung-Ang University, 1998

Submitted to the Graduate Faculty of

School of Information Sciences in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

University of Pittsburgh

2013 i

UNIVERSITY OF PITTSBURGH

SCHOOL OF INFORMATION SCIENCES

This dissertation was presented by Jong Do Park It was defended on December 14, 2012 and approved by Dr. Xia Lin, Professor, College of Information Science and Technology, Drexel University Dr. Leanne Bowler, Assistant Professor, School of Information Sciences Dr. Jung Sun Oh, Assistant Professor, School of Information Sciences Dissertation Advisor: Dr. Daqing He, Associate Professor, School of Information Sciences ii

AUTOMATED QUESTION TRIAGE FOR SOCIAL REFERENCE:

A STUDY OF ADOPTING DECISION FACTORS FROM DIGITAL REFERENCE

Jong Do Park, Ph.D.

University of Pittsburgh, 2013 Copyright © by Jong Do Park iii

AUTOMATED QUESTION TRIAGE FOR SOCIAL REFERENCE:

A STUDY OF ADOPTING DECISION FACTORS FROM DIGITAL REFERENCE

Jong Do Park, Ph.D.

University of Pittsburgh, 2013 The increasing popularity of Social Reference (SR) services has enabled a corresponding growth in the number of users engaging in them as well as in the number of questions submitted to the services. However, the efficiency and quality of the services are being challenged because a large quantity of the questions have not been answered or satisfied for quite a long time. In this dissertation project, I propose using expert finding techniques to construct an automated Question Triage (QT) approach to resolve this problem. QT has been established in Digital Reference (DR) for some time, but it is not available in SR. This means designing an automated QT mechanism for SR is very innovative.

In this project, I first examined important factors affecting triage decisions in DR, and extended this to the SR setting by investigating importantfactors affecting the decision making of QT in the SR setting. The study was conducted using question-answer pairs collected from Ask Metafilter, a popular SR site. For the evaluation, logistic regression analyses were conducted to examine which factors would significantly affect the performance of predicting relevant answerers to questions.

The study results showed that the user’s answering activity is the most important factor affecting the triage decision of SR, followed by the user’s general performance in providing good answers and the degree of their interest in the question topic. The proposed algorithm,

–  –  –

the performance of automated QT above the baseline for estimating relevant answerers to questions.

The results of the current study have important implications for research and practice in automated QT for SR. Furthermore, the results will offer insights into designing userparticipatory DR systems.

–  –  –

It would not have been possible to write this doctoral dissertation without the help and support of the kind people around me, to only some of whom it is possible to give particular mention here.

Above all, I would like to thank my wife Dr. Saem Han for her personal support and great patience at all times. My mother, parents-in-law, and brothers have gave me their unequivocal support throughout, as always, for which my mere expression of thanks likewise does not suffice.

I would like to express the deepest appreciation to my advisor, Dr. Daqing He, who has the attitude and the substance of a genius: he continually and convincingly conveyed a spirit of adventure in regard to research and scholarship. Without his patience, guidance and persistent help this dissertation would not have been possible.

I also would like to thank my committee members of Dr. Xia Lin, Dr. Leanne Bowler, and Dr. Jung Sun Oh for their support with constructive comments.

Lastly, I am most grateful to my beloved daughter, Joanna, for keeping me smile during the last few miles of my dissertation marathon race.

–  –  –

ACKNOWLEDGEMENTS

LIST OF TABLES

LIST OF FIGURES

1 INTRODUCTION

1.1 KEY CONCEPTS

1.1.1 Digital Reference (DR)

1.1.2 Expert and expertise

1.1.3 Expert finding

1.1.4 Reference

1.1.5 Social Reference (SR)

1.1.6 Question

1.1.7 Question Answering (QA)

1.1.8 Question Triage (QT)

1.1.9 Subject and topic

1.1.10 Element, attribute, and factor

1.2 BACKGROUND

–  –  –

librarians as answerers

1.2.3 DR: question asking and answering online

1.2.4 SR: user-participatory online reference

1.2.5 Challenges in SR services

1.2.6 The need for QT

1.3 PROBLEM STATEMENT

1.4 SCOPE OF THE STUDY





1.4.1 Focuses on human-mediated question answering in reference service..... 13 1.4.2 Focuses on automated QT

1.4.3 Focuses on expertise

1.4.4 Subject interest, performance, and contribution as expertise

1.4.5 Not system-building

1.5 RESEARCH GOAL

1.6 RESEARCH QUESTIONS

1.7 CONTRIBUTION OF THE STUDY

2 LITERATURE REVIEW

2.1 APPROACHES TO QA

2.1.1 Human-mediated QA

2.1.2 Approaches to automated QA

2.2 APPROACHES TO ACCESS EXPERTISE

2.2.1 Expertise seeking: the human-centered approach

2.2.2 Expertise retrieval: the system-centered approach

–  –  –

2.3.1 DR Model

2.3.2 QT

2.4 SR

2.4.1 The Social Web

2.4.2 What is SR?

2.4.3 SR Models

2.4.4 Three types of QA Sites

2.4.5 Approach to research on SR

2.5 OVERVIEW OF DIFFERENT APPROACHES TO QA

2.5.1 Key elements of the QA process

2.6 CHAPTER SUMMARY

3 RESEARCH DESIGN AND METHODOLOGY

3.1 INTRODUCTION

3.1.1 Purpose of the study

3.1.2 Research questions

3.1.3 Overview of the research design

3.1.4 Methodological challenges

3.2 PHASE 1: FACTOR IDENTIFICATION FOR QT FOR SR

3.2.1 Identification of the key elements of the QT process

3.2.2 Selection of attributes influencing QT

3.2.3 A framework of expertise: subject interest, performance, and contribution

–  –  –

3.3 PHASE 2: EXPERIMENT AND EVALUATION

3.3.1 Logistic regression analysis

3.3.2 Research steps in Phase 2

4 EXPERIMENTAL DESIGN

4.1 HYPOTHESES

4.1.1 Test level of topic interest (x1)

4.1.2 Test the scope of subject area of interest (x2)

4.1.3 Test the scope of topic area of interest (x3)

4.1.4 Test contribution to answer providing (x4)

4.1.5 Test Contribution to Question Submitting (x5)

4.1.6 Test performance to providing relevant answers (x6)

4.1.7 Test Response Time of Answer Providing (x7)

4.1.8 Test quota per day (x8)

4.2 A FRAMEWORK FOR AUTOMATED QT FOR SR

4.2.1 Question analysis module

4.2.2 Document retrieval module

4.2.3 Candidate recommendation module

4.3 PROCEDURES

4.3.1 Data preparation

4.3.2 Data processing

4.3.3 Candidate recommendation

4.3.4 Generating dataset for logistic regression analysis

–  –  –

5 RESULTS

5.1 PHASE 1: FACTOR IDENTIFICATION FOR QT FOR SR

5.1.1 Commonality and differences between DR and SR

5.1.2 Selecting attributes affecting to triage decision

5.1.3 Framework of user expertise

5.2 PHASE 2: EXPERIMENT AND EVALUATION

5.2.1 The logistic regression model

5.2.2 Imbalanced dataset

5.2.3 Data Normalization

5.2.4 Training

5.2.5 Selecting the best fit logistic model

6 DISCUSSION AND IMPLICATIONS

6.1 INTRODUCTION

6.2 THE AUTOMATED QT PROCESS

6.3 REPRESENTATION OF USER EXPERTISE

6.4 ATTRIBUTES THAT AFFECT THE TRIAGE PROCESS

6.4.1 Attributes of the question

6.4.2 Attributes of the answerer

6.4.3 Summary

6.5 IMPLICATIONS FOR RESEARCH DESIGN

6.5.1 Theoretical Implications

6.5.2 Practical implications

–  –  –

for SR?

7.1.3 Summary

7.2 LIMITATIONS OF THE STUDY

7.3 REMAINING QUESTIONS

7.4 FUTURE WORK

7.4.1 Selection of identifying additional factors affecting QT for SR............... 153 7.4.2 Question recommendation for SR sites

7.4.3 Study on automated question filtering

7.4.4 Study on user-participated library reference (SR)

BIBLIOGRAPHY

–  –  –

Table 1. Types of information systems and their approaches to QA

Table 2. Focuses for finding best answerer or experts

Table 3. Elements of QA process

Table 4. Fifteen factors affecting QT for DR (Pomerantz, 2003)

Table 5. Mapping the attribute of the question from DR into SR

Table 6. Mapping the attribute of the answerer from DR into SR

Table 7. Mapping the service attributes onto the answerer

Table 8. Mapping the triage factors of DR to SR

Table 9. Comparison of subject with topic

Table 10. The framework of expertise

Table 11. Research steps in phase 1

Table 12. Variables in the logistic regression analysis

Table 13. Research step in phase 2: building new models for QT in SR

Table 14. Summary of the dataset

Table 15. Summary of categories selected

Table 16. Summary of dataset statistics for training and testing

Table 17. Most frequent top 10 tags in 5 categories chosen.

–  –  –

Table 19. Example of the first four topics of computer-internet category using LDA model.

..... 97 Table 20. Variables used in the evaluation

Table 21. Logistic regression analysis of TAG-B0F1N20 dataset using a single factor for baseline

Table 22. Ratio between relevant cases and non-relevant case in the training dataset.

.............. 120 Table 23. Comparison of data type: raw, log, and z-transformation

Table 24. Performance comparison (MAP) of each model that used different set of training data with different conditions.

Table 25. Performance comparison between F0 and F1 models

Table 26. Performance comparison between the proposed model and baseline (training dataset)

Table 27. Performance comparison between the proposed model and baseline (testing dataset)

Table 28. Logistic regression analysis of factors affecting to the estimation of relevant answerers (Dataset of TAG-B0F0N20).

–  –  –

Figure 1. Conceptual framework of subject and topic

Figure 2. Overview of a generic QA architecture (Umbert, 2012)

Figure 3. General DR model (Pomerantz, Nicholson, Belanger, & David Lankes, 2004).

.......... 34 Figure 4. Model of answering behaviors in social Q&A: motivations & strategies (Oh, 2010).. 42 Figure 5. Conceptual frameworks of different approaches to QA

Figure 6. An overview of the research methodology

Figure 7. Proposed framework of QA process for SR

Figure 8. Map of user's subject areas of interest

Figure 9. Map of user's topic areas of interest

Figure 10. A process framework for automated question triage for social reference

Figure 11. Overview of the experiment procedure

Figure 12. Number of topics and perplexity in education category

Figure 13. Example of document-topic distribution: question 3905, question 3947

Figure 14. User-topic distribution (user 14648 and 1104)

Figure 15. Pseudo code for determining topics of the question (topic-clustering approach).

.... 104 Figure 16. Pseudo code for candidate selection - document-based approach

Figure 17. Raw distribution of answering activity

Figure 18. Z-distribution of answering activity

–  –  –

Information seeking is a common activity of human life. People sometimes rely on other human beings to solve their information problems. A familiar type of social interaction of information seeking is question asking. Question asking and answering are universal features of human communication (Goldman, 1999). The main purpose of asking questions is to learn the answer from the respondent to meet the questioner’s information need.

The development of Web 2.0 technologies, often referred to as the participatory Web, resulted in the growth of Social Reference (SR) services that enable users to interact with each other in the form of question asking and answering in online communities. The increasing popularity of SR services, in recent years, has enabled a corresponding growth in the number of users of SR services on the Web. This increase in the use of SR services has led to increases in the number of questions received by these services, thus the efficiency and quality of the services have become issues in this field.

In the field of libraries, reference librarians have practiced Question Triage (QT), or question routing, especially for Digital Reference (DR) to be able to handle increasingly large number of questions received, since the quality of the answers provided is directly affected by the amount of questions assigned to a reference service or an expert in DR services. Similarly, there is an obvious need to investigate QT for SR services, in order to increase efficiency and to improve the quality of the answers provided.



Pages:   || 2 | 3 | 4 | 5 |   ...   | 15 |


Similar works:

«WALKING THROUGH THE SHADOWS: RUINS, REFLECTIONS, AND RESISTANCE IN THE POSTCOLONIAL GOTHIC NOVEL A Dissertation Submitted to the School of Graduate Studies and Research in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy Sheri Ann Denison Indiana University of Pennsylvania May 2009 Indiana University of Pennsylvania The School of Graduate Studies and Research Department of English We hereby approve the dissertation of Sheri Ann Denison Candidate for the degree of...»

«THE INFLUENCE OF CULTURE ON CONSUMERS: EXPLORATORY AND RISK TAKING BEHAVIOUR A thesis submitted to the University of Minho in fulfilment of requirements for the degree of Doctor of Philosophy in Management Supervisors: Minoo Farhangmehr Aviv Shoham Acknowledgements This document is proof that this particularly difficult stage of my life is drawing to an end. Without the support of my family, supervisors, friends and colleagues this would not have been made possible. Looking back over this...»

«DIPLOMARBEIT Titel der Diplomarbeit “Die Comicreportage“ Verfasserin Judith Denkmayr Angestrebter akademischer Grad Magistra der Philosophie (Mag.phil.) Wien, im Juli 2008 Studienkennzahl lt. Studienblatt: A 301 295 Studienrichtung lt. Studienblatt: Publizistik und Kommunikationswissenschaft Betreuer: Ao. Univ.-Prof. Dr. Hannes Haas Ich versichere hiermit, dass ich die vorliegende Arbeit selbstständig verfasst, ganz oder in Teilen noch nicht als Prüfungsleistung vorgelegt und keine...»

«DISSERTATION: THE PRINCIPLE OF SOLIDARITY: A RESTATEMENT OF JOHN RAWLS´ LAW OF PEOPLES ZUR ERLANGUNG DES AKADEMISCHEN GRADES DOCTOR PHILOSOPHIAE (DR. PHIL) VON MILICA TRIFUNOVIĆ EINGEREICHT IM DEZEMBER 2011.AN DER PHILOSOPHISCHEN FAKULTÄT I DER HUMBOLDT-UNIVERSITÄT ZU BERLIN PRÄSIDENT DER HUMBOLDT-UNIVERSITÄT ZU BERLIN: PROF. DR. JAN-HENDRIK OLBERTZ DEKAN: PROF. MICHAEL SEADLE GUTACHTER: 1. PROF. DR. VOLKER GERHARDT 2. PROF. DR. WULF KELLERWESSEL TAG DER MÜNDLICHEN PRÜFUNG: 20. JUNI...»

«University of Alberta The Language and Literacy Practices of English-Chinese Bilingual Students in Western Canada by Miao Sun A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Elementary Education ©Miao Sun Fall 2011 Edmonton, Alberta Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for...»

«Grief, Gender and Mourning in Medieval North Atlantic Literature by Kristen Mills A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Centre for Medieval Studies University of Toronto © Copyright by Kristen Mills, 2013 Grief, Gender, and Mourning in Medieval North Atlantic Literature Kristen Mills Doctor of Philosophy Centre for Medieval Studies University of Toronto Abstract This dissertation explores the relationship between grief, cultural...»

«Florida State University Libraries Electronic Theses, Treatises and Dissertations The Graduate School Family Dynamics in American Literature: Genesis and Beyond Heidi Ann Marshall Follow this and additional works at the FSU Digital Library. For more information, please contact lib-ir@fsu.edu FLORIDA STATE UNIVERISTY COLLEGE OF ARTS AND SCIENCES FAMILY DYNAMICS IN AMERICAN LITERATURE: GENESIS AND BEYOND By HEIDI ANN MARSHALL A Dissertation submitted to the Department of English in partial...»

«Autorenliste Liste lateinischer Autoren und anonymer Werke des 13. Jahrhunderts (ca. 1170-1320) herausgegeben von Peter Christian JACOBSEN Redaktion: Martin Schieber M.A. (bis 1998) unter Mitarbeit von Elke Haag M.A., Dr. Georg Heldmann, Barbara Krehan, Tina Müller M.A., Sabine Walther M.A. und anderen Erlangen 2002 Die Datensammlung AUTOREN A ABSALON abbas SPRINGIERSBACENSIS (Abt 1190/96; † 1203): Werke: Sermones (Serm). Lit.: BibH; Fabr.; Gröb. II 197; LMA; Schneyer I 35-9. ABSTRACTUM: um...»

«Contact: home: 1701 E. Colter St. #390, Phoenix, AZ 85016; work: (480) 727-2170; cell: (865) 748-3837; email: chardgrove@gmail.com ! Education: Doctor of Philosophy, Geology, University of Tennessee (2004-2011) Bachelor of Science, Physics, Georgia Institute of Technology (1999-2004) ! Website: http://www.astrumterra.com/chardgrove Employment: Postdoctoral Research Scientist, Arizona State University (2013-present) Assistant Staff Scientist, Malin Space Science Systems (2012-2013) Postdoctoral...»

«enforce people’s contribution to the general good, as everyone naturally wants to do productive work, if they can find something they enjoy. Many communist anarchists believe that human behaviour is motivated by a combination of egoism and sympathy. Society develops towards greater sympathy and more refined moral ideals through a confrontation between ideals and reality. Some anarchists added an evolutionary argument, viz. that when groups compete, the more co-operative groups survive. There...»

«CURRICULUM VITAE Personal Data: Name: Abdul-Rahman Al-Ali Rank: Professor, Computer science and engineering dept. Address: American University of Sharjah P.O.Box 26666, Sharjah, UAE Phone: 971-6-515-2939 (Office) E-mail: aali@aus.edu URL : http://www.aus.edu/engr/coe/people/Al-Ali.php Education: May 1990: Doctor of Philosophy, Electrical Engineering Department, Vanderbilt University, Nashville, TN., U.S.A Minor: Computer science.. Ph.D. Thesis Title: On the Theory of High Sensitivity Metal...»

«Variation, Change and the Usage-based Approach Lynn Clark Thesis submitted for the degree of Doctor of Philosophy Linguistics and English Language The University of Edinburgh Abstract The potential for synthesis between variationist sociolinguistics and theoretical linguistics has been recognised by researchers in both sub-disciplines (e.g. Henry 1995; Adger and Smith 2005) but it has been difficult to move beyond a description of this unified approach towards an account of variation that can...»





 
<<  HOME   |    CONTACTS
2016 www.abstract.xlibx.info - Free e-library - Abstract, dissertation, book

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.