Berners-Lee, inventor of the World Wide Web (WWW), has a vision that the Web will not only provide human-human communication, but that machines will also be able to help. This vision is referred to as the semantic Web, because the meaning of the data on the Web will be available for machine processing. The authors of this book have written it as an introductory textbook, or self-study resource, providing an overview of the main ideas and techniques for the semantic Web. The book’s companion Web site (http://www.ics.forth.gr/isl/swprimer/) provides supplemental information, including presentation slides for each chapter. While this book does a good job of covering topics that surround the semantic Web, it is not an easy read. Readers need to be comfortable with predicate logic notations and concepts to be able to follow some of the more detailed sections of the book.

Chapter 1 takes a look at the Web, including how information is retrieved and some of the limitations of today’s tools. It then presents a scenario illustrating functions that may be available using semantic Web technologies. In the scenario, a personal agent is used to gather information and make recommendations on appointment times with physical therapists available through the user’s health insurance company. The personal agent will make use of key technologies, including metadata, ontologies, and logic.

The next few chapters discuss technologies available to provide data on data (metadata). They begin with the simplest, Extensible Markup Language (XML), and then discuss resource description framework (RDF), Web ontology language (OWL), and rules (logic and inference).

XML was developed to overcome some of the shortcomings of Hypertext Markup Language (HTML). XML is a metalanguage, allowing users to define their own tags, with few restrictions. It is the current de facto standard for supporting the machine processing of information. An XML document is said to be well-formed if it is syntactically correct. One of the syntax rules is that the root element--the outermost element--must be unique. The document can contain processing instructions for applications, with information on how to handle elements.

The structure of an XML document is maintained using schemas or document type definitions (DTDs). It can be formatted using Extensible Stylesheet Language (XSL), and transformed into another structure using XSL Transformations (XSLT). All of this together provides support for the exchange of information between different applications. However, using XML, machines are not able to understand the meaning of the information they are processing.

Resource description framework (RDF) is a data model with syntax provided by XML or other syntactic representations. RDF Schema (RDFS) defines the vocabulary of an RDF data model, making it machine accessible. RDF uses an object-attribute-value triple, also called a statement, to describe resources and their properties. Properties are used to describe relations between resources, and are themselves resources. These properties can be defined using uniform resource identifiers (URIs), which are usually uniform resource locators (URLs). RDF uses a reification mechanism allowing statements about statements. This requires that an auxiliary object be related to the original statement through the subject, predicate, and object properties. RDFS is considered a primitive ontology language.

OWL is designed to be the accepted ontology language of the semantic Web. Ontology languages are used to write explicit, formal conceptualizations of domain models. Mapping OWL on logics provides formal semantics and reasoning support. Reasoning support can be used to verify that the ontology and knowledge are consistent. OWL is defined as three different sublanguages: OWL Full, which uses all OWL primitives; OWL description logic (DL); and OWL Lite. OWL mainly uses RDF and RDFS. OWL is being used today, and is being extended to provide additional logical features such as rules.

In the final chapter, the authors analyze the current state of semantic Web research. According to them, many of the controversies stem from different interpretations of the semantic Web, one being that the semantic Web is a Web of data, and another that it is an enrichment of the current Web.

One of the criticisms of the semantic Web is that there is too much emphasis on the semantic and not enough emphasis on the Web. This was true early on, but recently there has been more emphasis on Web aspects. An example is the deployment of friend-of-a-friend (FOAF) technology. FOAF technology uses XML and RDF to mark up information. It appears to be similar to microformats, which have their own markup defined.

Appendix A lists the abstract OWL syntax for the code in chapter 4.

This book is a fairly technical overview of semantic Web technologies. It should meet the needs of its intended audience in an academic environment. To be able to apply the information in this book, readers will need to reference resources listed at the end of each chapter and on the companion Web site.