The apparent simplicity of the definition of metadata – data about data or information about information – actually neatly encompasses its breadth and complexity.
Metadata is a fundamental component of the most cutting edge business applications, but it is also a centuries old archival and bibliographic tool. Its scope can be vast or granular. It can describe one small resource, or all possible entities and their interrelationships in either a local or universal context. It can stand alone, or require integration with multiple distributed systems. It can serve an immediate business purpose or convey meaning and understanding through time. It can be proprietary and hardwired, or open source and hand crafted. It is a single concept that is owned, interpreted and codified by hundreds of different communities, that serves thousands of purposes and is implemented in millions of diverse paper and digital applications. And, in our complex digital world, it is increasingly the sole means that digital information resources can be accessed, managed, understood and maintained both immediately and through time.
There is no denying then, that metadata is hard. And because metadata as a concept is necessarily “owned” by so many different communities, it has always been a challenge to find basic primers and expert advice that is relevant to anyone coming fresh to this concept, or wanting to build on their current knowledge. But now, if someone does need to gain an understanding of this fluid, scalable, contested, important and complex concept, I would recommend that they start with this book.
Stephen Miller has done an exemplary job. He is clearly fantastically knowledgeable about the subject matter. Miller has years of practical experience in metadata implementation, Dublin Core development, as well extensive experience in university‐based metadata research and teaching.
With Metadata for Digital Collections, Miller has written a metadata guide that is clear and easy to understand. Although designed to support metadata initiatives in digital library, archival and museum communities (and while perfect for students and practitioners in these communities) the book, I believe, also has a much broader relevance. This is because, while its extensive examples are derived from library and archival implementations, the book itself goes back to first principles. It describes metadata, metadata schemes, interoperability, mapping, encoding, usability and more, in clear and comprehensible ways. Building from first principles the book then extends each concept so that the reader is left with a real working understanding of both the concepts and their implementation. This understanding can then be applied to a range of metadata projects, from current business through to cultural implementations.
Presented clearly and simply, Miller's work really does understand the totality of metadata and explains it as both an abstract concept and a technical reality. Today many other how‐to guides emphasise the technical basis of metadata and use technological language and notation. These can be challenging to both read and understand. Guides of this type can also tend to downplay the equally (more?) important theoretical or conceptual components of metadata. Miller, in contrast, understands and really defines metadata as a process. His book highlights the tight integration necessary between metadata and purpose, and the book cleverly takes readers through the variety of components in an effective metadata strategy. Good metadata is a matter of design, not just technology, and Miller really makes this clear. And all his lessons are presented so well that you don't even necessarily realise the complexity of the concepts and processes that are being explained to you.
In terms of content, in the book Miller uses three of the most commonly used metadata schemes in digital collection communities – Dublin Core, the XML based MODS (Metadata Object Description Schema) and VRA Core (Visual Resources Association Core) – as tools to explore a host of issues around metadata implementation. With the application of this schema as examples, Miller is able to examine such issues as the design and documentation of metadata schema, metadata granularity, controlled vocabularies, content guidelines and metadata interoperability and quality. In the course of his explanations Miller also gives an excellent introduction to XML, linked data and the semantic web.
The significant amounts of content in this book are all wonderfully accessible thanks to its detailed table of contents and nine page index. You can therefore easily dip into the content to resolve specific issues, or read the book from cover to cover for a very comprehensive overview of metadata implementation. The book also contains over 150 excellent tables and illustrations. Many of these are real examples of metadata in implementation, derived from libraries and archives from across the USA. There is also very comprehensive supporting material available on a companion web site.
Miller's book is really a fantastic primer and one, I think, that will have an enduring relevance. Because of their direct relationship to technology, books about metadata can be seen as a time‐bound and likely to become redundant when the technology changes around them. However, I do not think this is a risk with Miller's book. Metadata is growing, not diminishing in professional relevance, and all the time it is becoming more and more complex to understand and apply. By identifying metadata as a process, explaining its individual components and defining the standards at play in its implementation, Miller has created a book that moves beyond an implementation context and provides core general principles that will have an ongoing professional relevance.
Metadata is definitely hard, but it is also fundamental. If it is not designed and deployed with thought and foresight, the information resources it controls and describes will be compromised, perhaps irrevocably. So metadata is something collectively that we need to get right. In the contemporary business environment, metadata is often regarded as temporary and facilitative, with its values to be shoehorned into pre‐existing data structures. Instead, in all implementation environments, metadata does needs to be seen as the valuable and enduring record that it is.
If you have ever struggled with metadata, I would whole‐heartedly recommend this book to you. Even if you haven't, I think many, many practitioners will benefit from the clear and concrete understandings in Miller's book. It is very clear that significant amounts of time and expertise have been invested in this book. As a community I think we should be incredibly grateful to Miller and to Facet Publishing for the wonderful scholarly yet accessible resource they have developed for us.
