Preference Query Analysis and Optimization

118 Pages·2017·6.352 MB·English

by  Yunjun Gao,Qing Liu (auth.)

SPRINGER BRIEFS IN COMPUTER SCIENCE
Yunjun Gao
Qing Liu
Preference Query
Analysis and
Optimization
123
SpringerBriefs in Computer Science
Series editors
Stan Zdonik, Brown University, Providence, Rhode Island, USA
Shashi Shekhar, University of Minnesota, Minneapolis, Minnesota, USA
Xindong Wu, University of Vermont, Burlington, Vermont, USA
LakhmiC.Jain,UniversityofSouthAustralia,Adelaide,SouthAustralia,Australia
David Padua, University of Illinois Urbana-Champaign, Urbana, Illinois, USA
Xuemin (Sherman) Shen, University of Waterloo, Waterloo, Ontario, Canada
Borko Furht, Florida Atlantic University, Boca Raton, Florida, USA
V.S. Subrahmanian, University of Maryland, College Park, Maryland, USA
Martial Hebert, Carnegie Mellon University, Pittsburgh, Pennsylvania, USA
Katsushi Ikeuchi, University of Tokyo, Tokyo, Japan
Bruno Siciliano, Università di Napoli Federico II, Napoli, Italy
Sushil Jajodia, George Mason University, Fairfax, Virginia, USA
Newton Lee, Newton Lee Laboratories, LLC, Tujunga, California, USA
More information about this series at http://www.springer.com/series/10028
Yunjun Gao Qing Liu
(cid:129)
Preference Query Analysis
and Optimization
123
YunjunGao QingLiu
ZhejiangUniversity ZhejiangUniversity
Hangzhou, Zhejiang Hangzhou, Zhejiang
China China
ISSN 2191-5768 ISSN 2191-5776 (electronic)
SpringerBriefs inComputer Science
ISBN978-981-10-6634-4 ISBN978-981-10-6635-1 (eBook)
https://doi.org/10.1007/978-981-10-6635-1
LibraryofCongressControlNumber:2017956738
©TheAuthor(s)2017
This book was advertised with a copyright holder in the name of the publisher in error, whereas the
author(s)holdsthecopyright.
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
orinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar
methodologynowknownorhereafterdeveloped.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfrom
therelevantprotectivelawsandregulationsandthereforefreeforgeneraluse.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authorsortheeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinor
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations.
Printedonacid-freepaper
ThisSpringerimprintispublishedbySpringerNature
TheregisteredcompanyisSpringerNatureSingaporePteLtd.
Theregisteredcompanyaddressis:152BeachRoad,#21-01/04GatewayEast,Singapore189721,Singapore
Preface
The capability of database has been significantly improved over the decades.
However, the usability of database is far from meeting users’ requirements. This
book covers a comprehensive overview of preference query analysis and opti-
mization, which is the key tactic for improving database usability. It is worth
mentioning that this book spotlights on two representative preference queries, i.e.,
the reverse top-k query and the reverse skyline query.
The three key problems of preference query result analytics include: causality
and responsibility problem, why-not and why questions, and why-few and
why-manyquestions.Thisbookelaboratesfouraspectsofpreferencequeryanalysis
and optimization, including: (i) causality and responsibility problem: If preference
query results include users’ unexpected objects or users' expected objects do not
appearinthepreferencequeryresults,theusersmaywanttoknowwhatcausesthe
appearance of the unexpected objects and/or what causes the absence of the
expectedobjects.Towardthis,weexplorethecausalityandresponsibility problem
on probabilistic reverse skyline queries. (ii) Why-not and why questions: Usually,
the users would also like to know how to obtain expected preference query result
objects and/or exclude unexpected preference query result objects. In view of this,
we investigate the why-not and why questions on reverse top-k queries.
(iii) Why-few and why-many questions: In real applications, preference queries
mightreturntoofew(evenempty)ortoomanyanswerobjectstousers.Tothisend,
westudythewhy-fewandwhy-manyquestionsonreverseskylinequeries.(iv)We
developaninteractivesystem,termedasIS2R,toanalyzeunexpectedreversetop-k
queryresults.Basedonthefeedbacksfromusers,thesystemofferstheexplanations
of the unexpected reverse top-k query results as well as the suggestions on how to
gettheexpectedreversetop-kqueryresults,totheusers,including(a)tocontainthe
expected objects in reverse top-k query results, (b) to remove the unexpected
objects from reverse top-k query results, (c) to increase the reverse top-k query
result objects, and (d) to reduce the reverse top-k query result objects.
This book is aimed toward the readers with an interest in database usability. It
provides researchers and postgraduates a comprehensive overview of the general
conceptsandtechniquesforpreferencequeryanalysisandoptimization.Thisbook
v
vi Preface
can be an introductory book for the newcomers in the related research areas.
Moreover, it is also able to help practitioners and developers to improve the
usability of the database.
Hangzhou, Zhejiang, China Yunjun Gao
September 2017 Qing Liu
Acknowledgements
This work is partially supported by the 973 Program of China under Grant
No. 2015CB352502, the National Nature Science Foundation of China (NSFC)
under Grant No. 61522208 and 61379033, and the NSFC-Zhejiang Joint Fund
under Grant No. U1609217.
The authors conducted the research at the College of Computer Science,
ZhejiangUniversity,Hangzhou,China.Duringtheprocessofwritingthisbook,we
refer to the scientific researches of many scholars and experts. Consequently, we
would like to express our heartfelt thanks to them.
We thank Prof. Gang Chen from Zhejiang University (China), Associate
Professor Baihua Zheng from Singapore Management University (Singapore), and
Assistant Professor Lu Chen from Aalborg University (Denmark), for their
countless discussions and constructive comments. Meanwhile, we would like to
thank Miss. Linlin Zhou (with now working at NetEase (Hangzhou) Research
Institute) and Miss Wanqi Liu (with now working toward the Ph.D. degree at the
UniversityofTechnology,Sydney(Australia)), fortheireffectivealgorithmic code
implementation. Also, we want to thank Senior Editor Celine Lanlan Chang and
EditorJaneLifromSpringerforthevaluablefeedbacktoimprovethepresentation
of this book.
Lastbutnotleast,wewouldliketothankourownfamiliesforhavingourback.
Thanks a lot for all of your kind understanding and great support.
vii
Contents
1 Introduction to Preference Query Analysis and Optimization. . . . . . 1
1.1 Query Analysis and Optimization . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Preference Queries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Quantitative Preference Queries . . . . . . . . . . . . . . . . . . . . 3
1.2.2 Qualitative Preference Queries . . . . . . . . . . . . . . . . . . . . . 4
1.3 Research Issues and Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Overview of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Causality and Responsibility Problem on Probabilistic Reverse
Skyline Queries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Related Work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3.1 CRP Formulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3.2 CR2PRSQ Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 CP Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.6 Performance Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.6.1 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.6.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.7 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3 Why-Not and Why Questions on Reverse Top-k Queries . . . . . . . . . 31
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Related Work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.4 Answering Why-Not Questions. . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.4.1 Modifying Query Point . . . . . . . . . . . . . . . . . . . . . . . . . . 40
ix
x Contents
3.4.2 Modifying Why-Not Weighting Vector and k . . . . . . . . . . 45
3.4.3 Modifying Query Point, Why-Not Weighting Vector,
and k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.5 Answering Why Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.5.1 Modifying Query Point . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.5.2 Modifying Why Weighting Vector and k . . . . . . . . . . . . . 58
3.5.3 Modifying Query Point, Why Weighting Vector, and k . . . 61
3.6 Performance Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.6.1 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.6.2 Results on Why-Not Questions. . . . . . . . . . . . . . . . . . . . . 64
3.6.3 Results on Why Questions . . . . . . . . . . . . . . . . . . . . . . . . 69
3.7 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4 Why-Few and Why-Many Questions on Reverse
Skyline Queries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.2 Related Work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.3 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.4 Answering Why-Few and Why-Many Questions . . . . . . . . . . . . . 80
4.4.1 RI Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.4.2 SP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.5 Performance Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.5.1 Experimental Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.5.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.6 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5 Reverse Top-k Query Result Analysis and Refinement System. . . . . 101
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2 System Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.3 Demonstration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
6 Conclusion and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.2 Future Work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110