Knitting the Semantic Web

Knitting the Semantic Web explores the semantic web concept and what roles librarians and information professionals can play in producing a more “library‐like” web. As the editors note, the library community has largely ignored efforts to turn the web into something resembling the original vision of its inventor, Tim Berners‐Lee – a network where content can be expressed in natural language but marked‐up in such a way that software agents can understand and use that content for purposes other than merely formatting documents.

The first part of the book discusses the nature of the semantic web and introduces aspects of the semantic web of particular relevance to librarians: RDF (Resource Description Framework) and the use of URIs (Universal Resource Identifiers) to produce machine‐understandable descriptions of relationships between things, people, and concepts; SKOS (Simple Knowledge Organisation System) – a set of languages for representing structured, controlled vocabularies; and OWL (Web Ontology Language) for defining web‐based ontologies. Other chapters also discuss the application of Library of Congress controlled vocabularies to the semantic web, and possible roles for semantic web technologies in libraries.

The second part presents various semantic web projects, illustrating applications in biomedicine, physics and agriculture, and the use of the FOAF (Friend of a Friend) vocabulary specification in a real‐world situation. These papers clarify much of what is introduced earlier, providing examples of the actual thinking and planning that are required to implement the semantic web approach. Several papers show the sort of mark‐up of content involved; a paper by Liang et al. also includes an illustration of part of a metadata record expressed in RDF.

The book's origin as an issue of a scholarly journal does show. For instance, even in the synopsis above, the acronyms and jargon quickly pile up, and the inclusion of a glossary would have helped, despite the comprehensive index. Also, the first section could have used a more straightforward introductory chapter to orient newcomers to the subject matter and indicate how the various elements of the semantic web actually fit together. Symptoms of this include repeated references in several chapters to Berners‐Lee's Scientific American article, and the inclusion of the same diagram of the various semantic web technologies in three different chapters with no really useful explanatory text. In addition, despite the obvious expertise of the authors, the quality of writing varies greatly between chapters, something which cannot always be excused by the varying degrees of technical detail involved. Nevertheless, the editors' choice of chapter topics is excellent, always building their case that libraries and the semantic web are of mutual relevance; for instance, Greenberg's own chapter in the second half of the book describes the applicability of collection development, cataloguing, reference and circulation to semantic web development. It is difficult to know just how widely implemented a semantic version of the internet would be, but what it could bring to the exchange of data, information and knowledge even if limited to library, academic or commercial settings is undeniable. And that is why librarians should be paying attention.