Metadata for Digital Collections, Second Edition. By Steven Jack Miller. Chicago: ALA Neal-Schuman, 2022. 505 p. $69.99 softcover (ISBN 978-0-8389-4748-7).

The ability to create digital collections has become more accessible to libraries and cultural heritage institutions of all sizes over the last few years, making it a good time to release the second edition of Steven Jack Miller’s Metadata for Digital Collections. Miller is successful in his goal to write a textbook that is “accessible to beginners and library and information studies students as well as experienced professionals with little formal metadata training” (xxl). Throughout the book he introduces fundamental concepts that everyone developing a digital collection will need in order to be successful. Complex concepts—such as interoperability, linked data, and controlled vocabulary—are introduced early in the book in a manner that is appropriate for someone who is unfamiliar with the topics. Later chapters examine each of these topics in-depth providing a solid grounding and understanding of these concepts. In addition to writing plainly, Miller includes many examples that clearly illustrate the concepts being introduced. He also created a number of useful sidebars; I particularly found the “Typology of Metadata Standards” (15) and “Five Ways to Improve Metadata Quality and Interoperability: Summary Overview” (316) sidebars helpful in how they summarize the information in their respective sections which will make it easy for me to reference in the future.

Organized in a logical manner, the book starts with two chapters that introduce concepts that will be built upon throughout. These chapters also lay out a solid foundation to metadata, both in general and as it applies specifically to digital collections and describing the items in your digital collections. In the discussion of resource description, Miller includes information on best practices, but then describes what actually happens in many libraries that don’t have the staff or time to follow them. There is also a brief discussion on the potential value that users can add to the description of a resource and how best to incorporate that into your metadata. These are followed by a deep dive into Dublin Core, currently the most common content standard used in libraries, before moving on to chapters that delve into the specifics of resource description. These chapters are applicable across content standards but use Dublin Core in their examples. XML encoding, Metadata Object Description Schema (MODS), and Visual Resources Association (VRA) are introduced next; many libraries are starting to use MODS and VRA is common among museums and other cultural heritage institutions. The final three chapters address the importance of metadata interoperability, a brief introduction to linked data, and developing a metadata application profile. Miller does a good job of explaining linked data in an approachable manner that includes the realities of linked data in libraries. He explains that “there has been a great deal of hype about Linked Data, which can lead to unrealistic and over-inflated expectations. In the end, Linked Data is still just data” (390).

While I found the entire book useful, there were three parts that stood out: the three chapters on describing resources, the chapter on metadata interoperability and quality, and the final chapter on developing a metadata application profile. The first two chapters on describing resources cover the various kinds of information contained in a record: titles, dates, rights, formats, subjects, etc. In each section there is deep discussion of what kinds of information go in each part of a metadata record, different ways the metadata can be formatted, and real-life situations to consider when thinking about how to record the information. At the end of the first of these chapters Miller reminds us that “metadata scheme designers and metadata creators need to meet the resource discovery needs of their users” (119). This is something that I believe should be stressed in all cataloging—not just in digital collections—and should be incorporated more in literature and practice. In the section on subjects, not only does Miller discuss the coding of subjects and where in a metadata record they go, he also gives a thorough grounding on subject analysis, exhaustivity, specificity, and how it all relates to indexing. One of the strongest parts of this is his discussion of remaining objective and that metadata creators need “to avoid projecting their own subjective interpretations onto an image and into its metadata” (141). Finally, the third chapter on describing resources is an in-depth introduction to controlled vocabularies. This is particularly important when many people and systems want to rely on keyword searching, but Miller provides great examples on why keyword searching isn’t perfect. He also provides a brief explanation in this chapter on when and why a library might want to create their own controlled vocabulary for a project.

As technology has advanced and more libraries are starting to make their digital collections more accessible online, it has become clearer that metadata about the collections needs to be able to be shared. Even if a library doesn’t share their collection widely online, they still need to think about interoperability of their metadata. Due to continually evolving systems and technology, a library might migrate a collection to new or different platforms multiple times. By introducing the idea at the beginning of the learning cycle, whoever is planning the collection will be able to think and put plans in place. Miller makes an excellent argument that part of what makes sharing (or having your metadata harvested) successful is having quality data. As such, a good portion of chapter 10 “Metadata Interoperability, Shareability, and Quality” is on creating and maintaining quality metadata. One of my favorite sidebars is in this chapter (“Five Ways to Improve Metadata Quality and Interoperability: Summary Overview” (316)) and Miller goes into depth on all of the five suggestions in this chapter, including my favorite: documenting local practices.

Another highlight of the book for me is the final chapter, “Metadata Application Profile Design.” It brings together all of the concepts introduced in the preceding chapters and brings them into the day-to-day work that someone creating a digital collection will need to do. This provides a chance for a student to create something to apply what they’ve learned or for a library to create the local standard which they will implement when they start building their digital collection.

Overall, the new edition of Metadata for Digital Collections is a strong introduction to describing a library’s digital collection. It would make an excellent textbook for a class on metadata, as it approaches each topic sensibly, comprehensively, and is written to be understood by anyone who is not familiar with metadata. Even if a professor does not adopt the full textbook, there are several chapters that could be used to support student learning in a variety of courses, particularly the “Introduction to Metadata for Digital Collections,” “Controlled Vocabularies for Improved Resource Discovery,” and “Linked Data and Ontologies” chapters. In addition, this title would be valuable as part of a departmental reference collection, particularly at a small to medium-sized library that is starting to develop a digital collection, and where the librarians or staff need a good foundation in the concepts. Both students and practitioners can work their way through the book and come out at the end with a finished metadata application profile that could be implemented.—Lynn E. Gates (lgates@uccs.edu), University of Colorado Colorado Springs