Incorporating relational data into the Semantic Web [Elektronische Ressource] / vorgelegt von Cristian Pérez de Laborda Schwankhart

Incorporating relational data into the Semantic Web [Elektronische Ressource] / vorgelegt von Cristian Pérez de Laborda Schwankhart

-

English
157 Pages
Read
Download
Downloading requires you to have access to the YouScribe library
Learn all about the services we offer

Description

Incorporating Relational Data intothe Semantic WebInaugural-DissertationzurErlangung des Doktorgrades derMathematisch-Naturwissenschaftlichen Fakult¨atder Heinrich-Heine-Universit¨at Dus¨ seldorfvorgelegt vonCristian P´erez de Laborda Schwankhartaus Bilbao, SpanienAugust 2006Aus dem Institut fur¨ Informatikder Mathematisch-Naturwissenschaftlichen Fakult¨ atGedruckt mit der Genehmigung derMathematisch-Naturwissenschaftlichen Fakult¨at derHeinrich-Heine-Universit¨ at Duss¨ eldorfReferent: Prof. Dr. Stefan ConradKoreferent: Prof. Dr.-Ing. Kai-Uwe Sattler (TU Ilmenau)Tag der mundlic¨ hen Prufung:¨ 21.11.2006“[The Semantic Web] is about the data which currently is in rela-tional databases, XML documents, spreadsheets, and proprietary for-mat data files, and all of which would be useful to have access to asone huge database.” [Upd05]Tim Berners-LeeDirector of the W3CAcknowledgementsThis thesis is the result of three and a half years of work as a research and teachingassistant at the Databases and Information Systems group at the University ofDusseldorf.¨ I have to express my gratitude to all the people who made this workpossible.First of all, I like to express my warmest thanks to my supervisor Prof. Dr.Stefan Conrad for giving me the opportunity to join his research group and forencouraging me to realize my ideas on the Semantic Web. I also like to thankthe second reviewer of this thesis Prof. Dr.-Ing.

Subjects

Informations

Published by
Published 01 January 2006
Reads 19
Language English
Document size 1 MB
Report a problem

Incorporating Relational Data into
the Semantic Web
Inaugural-Dissertation
zur
Erlangung des Doktorgrades der
Mathematisch-Naturwissenschaftlichen Fakult¨at
der Heinrich-Heine-Universit¨at Dus¨ seldorf
vorgelegt von
Cristian P´erez de Laborda Schwankhart
aus Bilbao, Spanien
August 2006Aus dem Institut fur¨ Informatik
der Mathematisch-Naturwissenschaftlichen Fakult¨ at
Gedruckt mit der Genehmigung der
Mathematisch-Naturwissenschaftlichen Fakult¨at der
Heinrich-Heine-Universit¨ at Duss¨ eldorf
Referent: Prof. Dr. Stefan Conrad
Koreferent: Prof. Dr.-Ing. Kai-Uwe Sattler (TU Ilmenau)
Tag der mundlic¨ hen Prufung:¨ 21.11.2006“[The Semantic Web] is about the data which currently is in rela-
tional databases, XML documents, spreadsheets, and proprietary for-
mat data files, and all of which would be useful to have access to as
one huge database.” [Upd05]
Tim Berners-Lee
Director of the W3CAcknowledgements
This thesis is the result of three and a half years of work as a research and teaching
assistant at the Databases and Information Systems group at the University of
Dusseldorf.¨ I have to express my gratitude to all the people who made this work
possible.
First of all, I like to express my warmest thanks to my supervisor Prof. Dr.
Stefan Conrad for giving me the opportunity to join his research group and for
encouraging me to realize my ideas on the Semantic Web. I also like to thank
the second reviewer of this thesis Prof. Dr.-Ing. Kai-Uwe Sattler for his interest
in my work and the time for reading and commenting this thesis.
This work would not have been possible without the cooperation and support
of my colleagues and friends at the Institute of Computer Science, who created
a wonderful atmosphere. I thank Dr. Christopher Popfinger, my longtime fel-
low student for the joint studies, research, and activities performed during the
last years in Dus¨ seldorf and Munich. I also thank Dr. Evguenia Altareva, espe-
cially for her patience in sharing an office with me. Additionally, I extend my
compliments to Johanna Vompras, Tobias Riege, Christian Lochert, and all the
remaining people who helped me to have a great time here in Dusseldorf¨ doing
no research.
Guido K¨ onigstein and Marga Potthoff supported the realization of this the-
sis through their background help. I would probably have failed without them
because of technical or administrative barriers.
I am also grateful to all the students who helped to mature and realize my
ideas through their work and questions, especially to Yves Thielan, Matth¨ aus
Zloch, and Przemysla w Dzikowski.
Last but not least, I want to dedicate this thesis to my whole family,
especially to my parents, who continuously supported and encouraged me from
the very beginning. I would not have achieved anything without them.
Dusseldorf,¨ August 2006 Cristian P´erez de Laborda Schwankhartiv AcknowledgementsAbstract
The aim of the Semantic Web as promoted by the World Wide Web Consortium
is to enable software agents to access distributed information and to apply infer-
ence rules, so that new knowledge can be deduced. Nevertheless, since the vast
majority of data is still modeled and stored in relational databases, this infor-
mation is out of reach for most Semantic Web applications. Consequently, local
users usually create their own manual relational to semantic mappings or convert
it in a manual, time-consuming, and error-prone process into a corresponding
Semantic Web representation.
In this thesis we present Relational.OWL, a technique to automatically con-
vert a relational database into a Semantic Web representation, enabling Seman-
tic Web applications to access data actually stored in relational databases using
their own built-in functionality. Since the Relational.OWL representation of the
database does not result in objects containing real semantics, we additionally
show how to create mappings from the relational model to a target ontology
using an arbitrary closed RDF query language.
Using current Semantic Web techniques, a formerly relational data item, once
converted to its Semantic Web representation, can neither be identified unam-
biguously, nor be backtracked to its original storage location in the relational
database. We hence introduce the novel URI scheme db for identifying not only
databases, but also their schema and data components like tables or columns,
giving us the possibility to specify the exact and identifying storage location of
any data item in its original data source.
Since sharing information in distributed environments like the Semantic Web
often leads to recurring problems, we finally present the Link Pattern Catalog as a
modeling guideline for problems appearing during the design and implementation
of such information sharing environments.vi AbstractContents
1 Motivation 1
1.1 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Background 7
2.1 Relational Databases . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.1 Relational Model . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.1.1 Relational Data Model . . . . . . . . . . . . . . . 8
2.1.1.2 Relational Algebra . . . . . . . . . . . . . . . . . 9
2.1.2 SQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 The Semantic Web . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.2 Semantic Web Technologies . . . . . . . . . . . . . . . . . 17
2.2.2.1 RDF . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2.2.2 OWL Web Ontology Language . . . . . . . . . . 22
2.2.2.3 RDF Query Languages . . . . . . . . . . . . . . . 23
3 Bridging the Semantic Gap 27
3.1 Relational.OWL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.1.1.1 Data for the Semantic Web . . . . . . . . . . . . 29
3.1.1.2 Relational.OWL as an Exchange Format . . . . . 30
3.1.1.3 Peer-to-Peer Databases . . . . . . . . . . . . . . 32
3.1.2 Relevant Metadata . . . . . . . . . . . . . . . . . . . . . . 34
3.1.3 The Relational.OWL Ontology . . . . . . . . . . . . . . . 35
3.1.4 Schema Representation . . . . . . . . . . . . . . . . . . . . 37
3.1.5 Datantation . . . . . . . . . . . . . . . . . . . . . 38
3.1.6 Data Overhead in the Data Exchange Process . . . . . . . 40
3.2 Relational.OWL and RDF Query Languages . . . . . . . . . . . . 41
3.2.1 Relational.OWL and RDQL . . . . . . . . . . . . . . . . . 42
3.2.1.1 Selection . . . . . . . . . . . . . . . . . . . . . . 43
3.2.1.2 Projection . . . . . . . . . . . . . . . . . . . . . . 44
3.2.1.3 Set Union . . . . . . . . . . . . . . . . . . . . . . 45viii CONTENTS
3.2.1.4 Set Difference . . . . . . . . . . . . . . . . . . . . 45
3.2.1.5 Cartesian Product . . . . . . . . . . . . . . . . . 46
3.2.1.6 (Equi-)Join . . . . . . . . . . . . . . . . . . . . . 47
3.2.1.7 Discussion . . . . . . . . . . . . . . . . . . . . . . 48
3.2.2 Relational.OWL and SPARQL . . . . . . . . . . . . . . . . 49
3.2.2.1 Selection . . . . . . . . . . . . . . . . . . . . . . 49
3.2.2.2 Projection . . . . . . . . . . . . . . . . . . . . . . 50
3.2.2.3 Set Union . . . . . . . . . . . . . . . . . . . . . . 51
3.2.2.4 Set Difference . . . . . . . . . . . . . . . . . . . . 52
3.2.2.5 Cartesian Product . . . . . . . . . . . . . . . . . 53
3.2.2.6 (Equi-)Join . . . . . . . . . . . . . . . . . . . . . 54
3.2.2.7 Discussion . . . . . . . . . . . . . . . . . . . . . . 55
3.3 Relational to Semantic Mapping . . . . . . . . . . . . . . . . . . . 56
3.3.1 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.3.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.3.3 Mapping Process . . . . . . . . . . . . . . . . . . . . . . . 58
3.3.4 Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3.5 Classification . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.3.6 Sample Mapping . . . . . . . . . . . . . . . . . . . . . . . 61
3.3.7 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.4 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.5 Discussion and Future Work . . . . . . . . . . . . . . . . . . . . . 65
4 A Novel URI for Databases 67
4.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.2 Challenges Designing an Identifier . . . . . . . . . . . . . . . . . . 69
4.3 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.4 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.4.1 XML Data Exchange . . . . . . . . . . . . . . . . . . . . . 73
4.4.2 Relational.OWL Representation of a Database . . . . . . . 74
4.5 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.6 Discussion and Future Work . . . . . . . . . . . . . . . . . . . . . 77
5 Applications 79
5.1 Relational.OWL Implementations . . . . . . . . . . . . . . . . . . 79
5.1.1 Relational.OWL Application . . . . . . . . . . . . . . . . . 80
5.1.1.1 Introduction . . . . . . . . . . . . . . . . . . . . 80
5.1.1.2 Usage . . . . . . . . . . . . . . . . . . . . . . . . 81
5.1.2 Relational.OWL with XSLT and XQuery . . . . . . . . . . 84
5.1.2.1 Introduction . . . . . . . . . . . . . . . . . . . . 84
5.1.2.2 Implementation . . . . . . . . . . . . . . . . . . . 85
5.2 RDQuery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 86