Metadata for Information Management and Retrieval by David Haynes, Facet Publishing, London, 2004, xiv + 186 pages, hardback, ISBN 1 85604 489 0 ; Price £39.95
Anyone scaling the heights of metadata for the management and retrieval of digital information for the first time can be forgiven a degree of initial bewilderment. The same goes for this article, so a glossary of terms found here are offered in the spirit of saving readers' time . David Haynes' book appears to go a long way to guiding its explorers through the foothills and beyond in this complete introduction to the subject for the information professional.
At first glance this work covers all the classic metadata basics covered by Caplan's Metadata Fundamentals for All Librarians . But Haynes moves away from the solitary library perspective and also looks at the relevance of metadata to digital information resources and recent progress in the field of metadata standards in much more detail.
The book is divided into three sections. The introductory part gets the ball rolling with an interesting preamble on the parallel evolutionary histories of metadata; which in turn explains why there are complications in establishing a concrete definition. It is at this point that Haynes explains how the book is structured. He merges two models (Michael Day's seven item model and Gilliland-Swetland's more specific classification ) to come up with a five-point model of the purposes of metadata. It is on this model that he bases the five chapters which make up the lion's share of the work.
The next two chapters give a brief overview of ways to express metadata (such as DTDs, SGML, XML) and recent attempts to work within a common framework using data models and standards. A brief explanation: data models (such as RDF, FRBR and OAIS) are representations of the objects, concepts and entities that exist while standards (such as DC, MARC, ONIX and IEEE LOM) are specific functional applications of metadata.
The second section of the book looks in more detail at the five-point model of the purposes of metadata previously mentioned. These purposes are:
Each chapter catalogues the various metadata attributes that would be applicable to the given purpose. Throughout his exploration Haynes always includes a digital resource slant. For example he spends significant time on Digital Object Identifiers (DOIs); and when looking at information retrieval, as well as subject indexing and controlled vocabularies, he also considers retrieval on the Internet, the algorithms used and discovery of newer resources such as multimedia and graphic images. Chapter eight on Interoperability and e-commerce initially feels a little awkward as the pearing is not one I would naturally plump for. However again Haynes has useful things to say, though he does have a tendency to be slightly e-government-centred. On reflection it is only natural for the author to make considerable mention of e-GIF as his recommendations to the Cabinet Office in 2000 led to the formation of the Working Group on Metadata and the eventual establishment of the e-Government Metadata Standard. I personally felt that the more significant interoperability issues, such as consistency, facilitating data exchange between systems etc. were covered in more detail in the Managing Metadata chapter.
The final section of the book looks at more general treatment of metadata and considers future developments. The chapter on Managing Metadata takes a project lifecycle approach and provides a good checklist for someone starting a new project. There is not a great deal on the creation of metadata but nevertheless Haynes makes a good stab at the issues involved when choosing schemas. Some important topics covered in this chapter include importing metadata and normalising data, controlled vocabularies and quality management.
In the final chapter Haynes takes a look at the potential for developing and using metadata concepts further. He feels that the increased interest in the study and use of metadata means that it has the capacity in future years to become an entirely new discipline. Important areas to watch are the semantic Web, what will happen to the cost of indexing and standards development. Haynes concludes that metadata is an integral part of information systems and that it will develop through co-operation between communities and the establishment of a universal model of metadata. XML schemas are likely to play a big part in this and indeed become the norm.
David Hayne's book is not, and does not claim to be, all things to all people. It is clearly not intended as a highly technical resource. However it does provide a very sound introduction to current metadata concepts for librarians and information workers alike. There is no doubt that it is already a key text on Library and Information Science courses and will remain so for many years.
This is a relatively slim publication at only 184 pages, but nonetheless a lot is covered in the space available. Many of topics and technologies are not dealt with in depth, but references are consistently given to enable further research. David Haynes is actually running a series of workshops that introduce the book and take some of the issues mentioned further .
The real success of this work lies in Haynes' ability to capture both the older issues of concern to librarians, archivists and museum and gallery professionals, such as what metadata to include, and the new information management puzzles that the Internet has generated. As such it should prove a valuable resource for many people.