Cur‚ / Blin RDF Database Systems

Triples Storage and SPARQL Query Processing
1. Auflage 2014
ISBN: 978-0-12-800470-8
Verlag: Elsevier Science & Techn.
Format: EPUB
Kopierschutz: 6 - ePub Watermark

E-Book, Englisch, 256 Seiten

RDF Database Systems is a cutting-edge guide that distills everything you need to know to effectively use or design an RDF database. This book starts with the basics of linked open data and covers the most recent research, practice, and technologies to help you leverage semantic technology. With an approach that combines technical detail with theoretical background, this book shows how to design and develop semantic web applications, data models, indexing and query processing solutions. - Understand the Semantic Web, RDF, RDFS, SPARQL, and OWL within the context of relational database management and NoSQL systems - Learn about the prevailing RDF triples solutions for both relational and non-relational databases, including column family, document, graph, and NoSQL - Implement systems using RDF data with helpful guidelines and various storage solutions for RDF - Process SPARQL queries with detailed explanations of query optimization, query plans, caching, and more - Evaluate which approaches and systems to use when developing Semantic Web applications with a helpful description of commercial and open-source systems

Olivier Cur‚ is an associate professor of computer science at the Universit‚ Paris-Est in France and is researching at the CNRS LIGM lab. He holds a Ph.D. in Artificial Intelligence from the Universit‚ de Paris V, France and has published three book chapters, eight journal articles and more than 50 papers in international, peer-reviewed conferences in the fields of databases, semantic web and ontologies. Professor Cur‚ has organized workshops including Ambient Data Integration (ADI) at On the Move (OTM) conference in 2008, 2009 and 2010. He has received three cooperative research grants to work with the Database and Information System research team of Pr. Stefan Jablonski at the University of Bayreuth, Germany. In 2013, Professor Cur‚ received a grant for a France-Stanford collaboration to conduct research with Stanford's BioMedical Informatics Research (BMIR) laboratory.

Autoren/Hrsg.

Curé, Olivier

Blin, Guillaume

Inhaltsverzeichnis

1;Front Cover;1
2;Social Support, Life Events, and Depression;4
3;Copyright Page;5
4;Table of Contents;6
5;Contributors;12
6;Preface;14
7;Acknowledgments;16
8;Part I: IDENTYFYING BASIC ISSUES AND APPROACH;18
8.1;Chapter 1. Social Support in Epidemiological Perspective;20
8.1.1;..TRODUCTION;20
8.1.2;CONCEPTUALIZATON;26
8.1.3;MEASUREMENTS;27
8.1.4;CAUSAL MODELING;30
8.1.5;SPECIFICATIONS AND ELABORATONS;30
8.2;Chapter 2. Conceptualizing Social Support;34
8.2.1;INTRODUOTON;34
8.2.2;C..C...U.LI..TI..S OF SOCIAL SUPPORT;34
8.2.3;THE S.....TIC DEFINITION OF SOCIAL SUPPORT;35
8.2.4;FURTHER DISCUSSION OF THE SYNTHETIC DEFINITION;37
8.2.5;A THEORY OF SOCIAL RESOURCES AND SOCIAL SUPPORT;43
8.2.6;DISCUSSION;46
8.3;Chapter 3. Study Design and Data;48
8.3.1;THE ALBANY AREA HEALTH SURVEY;48
8.3.2;THE PRETEST;50
8.3.3;SAMPLING DESIGN;51
8.3.4;REPRESENTATIVENESS OF THE SAMPLE;53
8.3.5;THE INTERVIEW SCHEDULES;56
8.3.6;TIME FRAMES FOR QUESTIONS;62
8.3.7;THE INTERVIEWING STAFF;64
8.3.8;SUMMARY;64
9;Part II: MEASURING DEPRESSION, LIFE EVENTS, AND PSYCHOLOGICAL RESOURCES;66
9.1;Chapter 4. Measuring Depression: The GES-D Scale;68
9.1.1;INTRODUCTION;68
9.1.2;MEASURING DEPRESSION: MOOD, SYMPTOM, OR SYNDROME;69
9.1.3;THE CENTER FOR EPIDEMIOLOGICAL STUDIES DEPRESSION (CES-D) SCALE: HISTORY OF DEVELOPMENT;70
9.1.4;TYPES OF DEPRESSIVE SYMPTOMATOLOGY;72
9.1.5;RELIABILITY AND VALIDITY OF THE CES-D SCALE IN THE CURRENT STUDY: A COMPARISON WITH PREVIOUS WORK;74
9.1.6;RELATONSHIP OF THE CES-D SCALE OVER TIME;80
9.1.7;THE CES-D AND CLINICAL CASENESS;83
9.1.8;SUMMARY;85
9.2;Chapter 5. Measuring Life Events;88
9.2.1;INTRODUOTON;88
9.2.2;THE STRESS(OR) CONSTRUCT;88
9.2.3;ANALYTIC TASKS;89
9.2.4;OUR MEASUREMENT OF LIFE EVENTS;91
9.2.5;STATISTICAL DESCRIPTION OF LIFE-EVENTS SCALES;93
9.2.6;CONCLUSIONS;110
9.3;Chapter 6. Measuring Psychological Resources;114
9.3.1;INTRODUCTION;114
9.3.2;PRESENT OBJECTIVES;115
9.3.3;THE CONCEPT OF PERSONAL COMPETENCE;116
9.3.4;THE CONCEPT OF SELF-ESTEEM;119
9.3.5;OVER-TI.. C.RRELATIONS ;123
9.3.6;TESTING THE PROXY ISSUE;123
9.3.7;FACTOR ANALYSES OF PERSONAL COMPETENCE AND SELF-ESTEEM;125
9.3.8;SUMMARY;127
10;Part III: MEASURING SOCIAL SUPPORT;130
10.1;Chapter 7. Measuring Intimate Support: The Family and Confidant Relationships;134
10.1.1;INTRODUCTION;134
10.1.2;ANALYSIS OF THE MEDALIE-GOLDBOURT SCALE OF FAMILY RELATIONSHIPS;136
10.1.3;CONFIDANT SUPPORT: CONCEPTUALIUZATION AND MEASUREMENT;139
10.1.4;DISCUSSION;143
10.1.5;SUMMARY AND CONCLUSIONS;144
10.2;Chapter 8. Measuring the Instrumental and Expressive Functions of Social Support;146
10.2.1;SCALE DEVELOPMENT;146
10.2.2;TOTAL SCALE RELIABILITY AND VALIDITY;147
10.2.3;DETERMINING DIMENSIONS OF THE INSTRUMENTAL AND EXPRESSIVE ITEMS;147
10.2.4;FURTHER DEVELOPMENT: STRONG-TIE SUPPORT;162
10.2.5;RELIABILITY AND VALIDITY OF STRONG-TEE SUPPORT;164
10.2.6;CRITICISMS OF THE INSTRUMENTAL AND EXPRESSIVE SUPPORTIVE SCALES: AN EMPIRICAL TEST;164
10.2.7;SUMMARY AND IMPLICATONS;167
10.3;Chapter 9. Measuring Community and Network Support;170
10.3.1;INTRODUCTION;170
10.3.2;COMMUNITY SUPPORT;172
10.3.3;NETWORK SUPPORT;175
10.3.4;CONCLUDING REMARKS;187
11;Part IV: Constructing and Estimating Basic Models;188
11.1;Chapter 10. Modeling the Effects of Social Support;190
11.1.1;INTRODUCTION;190
11.1.2;MODELING THE EFFECTS OF SOCIAL SUPPORT;192
11.1.3;.ROPERTIES AND IMPLICATIONS OF THE MODELS;195
11.1.4;EVIDENCE FROM OTHER STUDIES;197
11.1.5;DATA AND THE ...LYTIC TECHNIQUE;204
11.1.6;THE ADDITIVE MODELS (MODELS 2, 3, and 4);207
11.1.7;CLASS A MODELS;207
11.1.8;CLASS . MODELS;211
11.1.9;CLASS C MODELS;213
11.1.10;THE INTERACTIVE MODELS;213
11.1.11;MODELS OF JOINT ADDITIVE AND INTERACTIVE EFFECTS;215
11.1.12;MODELS OF JOINT ADDITIVE AND INTERACTIVE EFFECTS;215
11.1.13;CONCLUSIONS;223
12;Part V: EXPLORING BASIC MODELS;228
12.1;Chapter 11. The Age Structure and the Stress Process;230
12.1.1;RELATIONSHIP BETWEEN AGE AND DEPRESSION;232
12.1.2;CONSTRUCTION OF AGE CATEGORIES;233
12.1.3;AGE-RELATED EFFECTS OF LIFE EVENTS AND SOCIAL SUPPORT ON DEPRESSION;238
12.1.4;FURTHER AGE-GROUP REFINEMENTS;241
12.1.5;SUMMARY AND IMPLICATIONS;241
12.2;Chapter 12. Sex, Marital Status, and Depression: The Role of Life Events and Social Support;248
12.2.1;GENDER, MARITAL STATUS, AND DEPRESSION: A REVIEW;248
12.2.2;THE CONFOUNDING ISSUE: MARITAL STATUS AS A STRESSOR OR AS A SOCIAL SUPPORT;252
12.2.3;A MODEL OF SEX, MARITAL STATUS, UFE EVENTS, SOCIAL SUPPORT, AND DEPRESSION;253
12.2.4;SEX, MARITAL STATUS, AND DEPRESSION;254
12.2.5;SEX, MARITAL STATUS, LIFE EVENTS, AND SOCIAL SUPPORT;255
12.2.6;INDEPENDENT EFFECTS OF LIFE EVENTS AND SOCIAL SUPPORT;257
12.2.7;JOINT EFFECTS, MEDIATING EFFECTS, AND ....RACTION EFFECTS;259
12.2.8;SUMMARY AND IMPLICATIONS;263
12.3;Chapter 13. Social Class and Depressive Symptomatology;266
12.3.1;INTRODUCTION;266
12.3.2;MALE-FEMALE CLASS DIFFERENCE IN VULNERABILITY;267
12.3.3;SOCIOECONOMIC CHARACTERISTICS OF MALES AND FEMALES;268
12.3.4;SOCIAL CLASS, LIFE EVENTS, SOCIAL SUPPORT, AND DEPRESSION;271
12.3.5;CLASS-ORIENTED EFFECTS OF LIFE EVENTS AND SOCIAL SUPPORT ON DEPRESSION;274
12.3.6;CLASS AND THE SUPPRESSING ROLE OF SOCIAL SUPPORT;279
12.3.7;SUMMARY;281
12.4;Chapter 14. Prior History of Illness in the Basic Model;284
12.4.1;PRIOR HlSTORY OF ILLNESS;284
12.4.2;THE EVENT-PRONENESS MODEL;285
12.4.3;PHYSICAL ILLNESS AND PSYCHOLOGICAL DISTRESS;287
12.4.4;MODELS TO BE EXAMINED;288
12.4.5;THE MEASURE OF ADVERSE PHYSICAL HEALTH;289
12.4.6;MODEL 1: THE EVENT-PRONENESS HYPOTHESIS (PHYSICAL ILLNESS AND LIFE EVENTS);290
12.4.7;MODEL 2: THE MODIFIED EVENT-PRONENESS HYPOTHESIS: PRIOR PHYSICAL ILLNESS, LIFE EVENTS, AND SUBSEQUENT PSYCHOLOGICAL SYMPTOMS;292
12.4.8;MODEL 3: THE BASIC MODEL (LIFE EVENTS, SOCIAL SUPPORT, AND DEPRESSION) WITH PRIOR ILLNESS;293
12.4.9;CONCLUSIONS;296
13;Part VI: EXAMINING ALTERNATIVE APPROACHES TO THE BASIC MODELS;298
13.1;Chapter 15. Gender of the Gonfidant and Depression;300
13.1.1;GENDER DIFFERENCES IN WELL-BEING;300
13.1.2;DATA AND MEASUREMENT;308
13.1.3;MARITAL CHANGE AND CONFIDANTS;312
13.1.4;MULTIVARIATE ANALYSIS;317
13.1.5;SUMMARY AND DISCUSSION;322
13.2;Chapter 16. Buffering the Impact of the Most Important Life Event;324
13.2.1;INTRODUCTON;324
13.2.2;PREREQUISITES AND ELEMENTS OF THE BUFFERING MODEL;325
13.2.3;DESIGN AND HYPOTHESES;327
13.2.4;THE MEASURES;330
13.2.5;EFFECT OF THE MOST IMPORTANT LIFE EVENT;332
13.2.6;BUFFERING EFFECTS OF STRONG TIES;334
13.2.7;DISCUSSION;342
13.3;Chapter 17. Epilogue: In Retrospect and Prospect;350
13.3.1;SUMMABY OF MAJOR FINDINGS;350
13.3.2;FUTURE RESEARCH AGENDA;353
13.3.3;TOWARD A THEORY OF THE INTERNAL STRUCTURE OF SOCIAL SUPPORT;357
14;References;360
15;Author Index;380
16;Subject Index;388

Leseproben

Chapter Two

Database Management Systems

Abstract

In this chapter, we present the main aspects and solutions of database management systems that have inspired or are currently influencing RDF stores. This ranges from systems based on the relational model to NoSQL and the recent NewSQL stores. It covers aspects such as storage solutions, efficient query processing through indexation, data and processing distribution and parallelism.

Keywords

database management system

relational model

NoSQL

Structured Query Language

Index

ACID

Distribution

The objective of this chapter is to provide some background knowledge on database management systems, a software the purpose of which is to define, create, manage, query, and update a database. We present certain aspects of systems that have been the most widely used in production for the last couple of decades. We also consider some trends that have emerged during the last few years. We limit this investigation to systems that are currently being used or have been in the past as the backend of existing Resource Description Framework (RDF) database management systems. Due to space limitations, we cannot provide a thorough presentation of these systems, therefore we concentrate on some peculiar characteristics that motivated their adoption in RDF data management and are main differentiators among these systems.

The first category of systems we consider is the relational database management system (RDBMS). It is a very popular family of systems that has dominated the database market for the last 30 years. We provide a short introduction on this topic for readers coming with a (Semantic) Web background, but a complete presentation is out of the scope of this book, (Elsmari, 2010) and (Kifer, 2005) are good introductions. Nevertheless, we define some notions and terms that are going to be used throughout this book. Then, we introduce concepts that are needed for understanding the particularities of the different RDF systems that are studied in Part 2 of this book. These concepts are generally presented in books investigating the internal description of RDBMSs, and are usually not considered in classic RDBMS books that concentrate more on how to model for and use these systems.

The second kind of systems we address corresponds to NoSQL systems. These systems only appeared a couple of years ago but already benefit from a high adoption rate in many IT infrastructures. They are used in many different application domains and are far from being limited to the issues of Web companies. This presentation concerns the four main families of NoSQL systems and also introduces some tools that are frequently associated with these systems, such as the popular parallel MapReduce processing framework.

Finally, we provide trails on the evolution of RDBMS and NoSQL systems. For instance, we present an introduction to some novel approaches in developing RDBMSs. Here, we are mainly concerned with implementing systems that leverage on the evolution of the hardware environment—for example, the availability of larger main memory spaces, the emergence of solid-state drives (SSDs), and the emergence of cloud-based systems. We also emphasize on the appearance of novel functionalities in NoSQL database systems. All these aspects may play a role in the evolution of RDF database management systems in the near future.

2.1. Technologies prevailing in the relational domain

2.1.1. Relational model

The relational model was introduced by Edgar F. Codd in the early 1970s (Codd, 1970) and is the foundation of RDBMSs. In this model, the first-class citizen is a specific structure named a relation that contains tuples (a.k.a. records). All the tuples in a relation have the same set of fields (a.k.a. attributes) where each field has a certain type, such as an integer, a date, or a string. This matches the definition of structured data presented in Chapter 1. The operations performed over this model are handled by an algebra that principally serves as a query language to retrieve data. A selection operation retrieves information from a single relation or a set of relations. In this last case, relations are generally joined over some common-type attributes. An important aspect of this algebra is that it produces an output that takes the form of a structure that is itself a relation. Thus, this approach enables the composition of relational queries one into another—that is, using the result of a query as the input of another one.

The concepts pertaining in the relational model are transposed into an RDBMS using the following term translation: relations, attributes, and tuples correspond respectively to tables, columns, and rows (but these terms can be used interchangeably to denote the same notions). In an RDBMS, the query language is named Structured Query Language (SQL), and it implements most of the operations available in the relational algebra but also provides additional ones. Many consider that SQL is a major reason for the success and dominance of this type of data management system. This is mainly due to its short learning curve and its expressivity which has been defined to support good computational complexity properties. In other words, a lot of practical questions can be expressed with few concepts and answered relatively rapidly even over large data sets. Moreover, the wide adoption in most existing RDBMS systems of a common, standardized subset of SQL is another major advantage. Therefore, it enables one user to easily switch from one RDBMS to another with relative ease, such as from MySQL to Oracle or the other way around. SQL also extends the relation algebra by update operations, which are the ability to delete, insert, or modify information.

For example, we consider an oversimplified blog application containing the following information. Blog entries are being written by users, themselves being characterized by an identifier, first and last names, and a gender. For each blog entry, we store the content of the entry (i.e., the text it contains), its storage date (i.e., the date at which it’s being stored in the database), the user who has produced the entry, and the category of the entry. A category corresponds to a subject area, such as sports, technology, or science. Finally, a subscription feature enables end-users to follow the blog entries of other users. Many solutions are available to represent the corresponding conceptual data model (e.g., an entity relationship (ER) diagram), but we have opted for a Unified Modeling Language (UML) notation using a class diagram, see Figure 2.1. Note that in this diagram, we consider that identifiers are explicit (e.g., user identifier), and we therefore do not display them in the figure.

Figure 2.1Data model for the blog use case using UML notation.

When the target database is an RDBMS, the conceptual model is translated into a relation schema. This is presented with some sample data in Figure 2.2. The schema contains four tables, namely User, Category, Blog, and Follows. The first three are direct translations from the classes proposed in Figure 2.1. The latter corresponds to the follows role the cardinalities of which are many-to-many (the two * symbols in Figure 2.1), meaning that a user can follow as many users as he or she wants and he or she can also be followed by an unrestricted number of users. The attributes forming this relation correspond to user identifiers from both the follower and the followed users.

Figure 2.2Sample data for the blog use case.

We can also see that some columns have been added to some relations. This corresponds to many-to-one relationships between entities (represented as 1 * associations between boxes in Figure 2.1). This is aimed to support joins between relations—for example, the CategoryId of the Blog relation enables joins with the id attribute of the Category relation. This approach implies column redundancy, which allows the definition of queries that may be useful when designing domain application. For example, the following query retrieves all the blogs belonging to the Science category and that have been written by persons followed by Mary:

The capitalized terms correspond to reserved words of SQL. This query contains three clauses—SELECT, FROM, and WHERE—which respectively retrieve some columns to be displayed in the result set, specify the tables needed for the query to execute (AS is used for the creation of aliases for table names, which induces easier reading of the query), and can define some filters and/or joins. This query requires three tables and the WHERE clause contains two filters (i.e., on the label and followerId columns using the LIKE and equality operators) and two joins (e.g., on the columns id of Category and...