Linked Data for Cultural Heritage, edited by Ed Jones and Michele Seikel. Chicago: ALA Editions, 2016. xvi, 134 p. illus. ISBN 978-0-8389-1439-7. $75.00.
Linked Data for Cultural Heritage offers a useful selection of essays focusing on linked data projects and concepts that have emerged in the cultural heritage domain. Though certainly not comprehensive, this book brings together six articles by practitioners well known in the linked data world, resulting into an easily accessible overview.
The introduction written by one of its editors, Ed Jones, sets the stage for the following collection by giving a general background on linked open data and the challenges that still need to be overcome.
The first chapter “Linked Open Data and the Cultural Heritage Landscape,” by Hilary K. Thorsen and M. Cristina Patuelli, then narrows the focus to linked open data implementations in the cultural heritage domain. Both authors are key participants in the “Linked Jazz” project. Linked Jazz focuses on extracting relationships between jazz musicians from oral history transcripts thereby allowing for the visualization of social networks that existed in the community. Besides providing an in-depth description of their own project, Patuelli and Thorsen also present a selection of well-known cultural heritage linked data implementations. This overview includes two cultural heritage data aggregations (Europeana and the Digital Public Library of America (DPLA)), three archival projects (Social Network and Archival Context Project (SNAC), Repository for Linked Open Archival Data (RELOAD), and the collection of the Jewish Documentation Center Foundation (CDEC)) as well as two museum collections (Amsterdam Museum and the Smithsonian Museum of American Art). The authors highlight the innovative aspects of all of these projects, but also the obstacles that had to be overcome.
“Making MARC Agnostic,” by Carl Stahmer, is a case study of transforming the MARC-based English Short Title Catalogue (ESTC) into linked data using a triplestore data model. Stahmer’s explanations allow the user to develop a deeper understanding of linked data concepts. The project identified a tension between traditional cataloging and the researcher’s desire to add their scholarly input to the resource descriptions. The solution offered by Stahmer in the section on “Social Cataloging” for the planned linked data-based ESTC21 interface and workflow is of particular interest and illustrates how linked data can help us solve problems in ways traditional library cataloging could not.
Stahmer’s essay also alludes to the importance of controlled vocabularies to linked open data. Allison Jai O’Dell expands on this in her article “Authority Control for the Web.” O’Dell convincingly argues that many of the traditional library practices employed for the creation of controlled vocabularies (both value vocabularies and library element sets) are transferable to linked data, but should be expanded to better take advantage of the opportunities that the Semantic Web has to offer. This would allow us to establish relationships between vocabularies and to take advantage of data previously hidden in authority records.
The fourth chapter exploresthe importance of controlled vocabularies in the context of science, technical, and medical publishing (STM). In “Linked Data Implications for Authority Control and Vocabularies,” Iker Huerga and Michael P. Lauruhn suggest that linked data, particularly controlled vocabularies, presented in a Semantic Web-friendly manner, could be used to support new ways of data and research sharing that go beyond the traditional STM publishing model. Huerga and Lauruhn then present an overview of key components necessary to express vocabularies as linked data, including Uniform Resource Identifiers (URIs), Resource Description Framework (RDF), Web Ontology Language (OWL), and SPARQL Protocol and RDF Query Language (SPARQL).
The last chapters are devoted to two current high-interest linked data developments in the library domain: OCLC’s Schema.org implementation and the development of BIBFRAME at the Library of Congress. In her article “A Division of Labor,” Carol Jean Godby explains the reasons for OCLC’s decision to focus their efforts on Schema.org and provides a detailed account on experimenting with the standard, including the role of the Schema Bib Extend Community Group and the BiblioGraph.net extension vocabulary and lessons learned so far. Particularly worthwhile reading is the section on “Modeling Beyond Published Monographs” in which Godby illustrates quite convincingly how badly served the area of audiovisual resources really is in current library cataloging (using the example of a recording of a ballet performance based on a fairy tale), and then suggesting an alternative model that could potentially be realized using Schema.org with extensions.
In the last chapter in this collection “BIBFRAME and Linked Data for Libraries,” Sally McCullum provides background on the development of BIBFRAME, including its goals, history, and data model. The readers should be aware that this chapter was written before the release of BIBFRAME 2.0, thus it describes the data model of the previous BIBFRAME version. However, much of the information is still valuable for an understanding of the current 2.0 data model. In addition, this chapter offers a quick introduction to RDF and other linked data developments at the Library of Congress, such as Metadata Authority Description Schema in RDF (MADS/RDF) and the many vocabularies and code lists that have already been transformed to linked open data and are available on id.loc.gov.
Overall, this short collection (just 134 pages) is well worth reading–certainly for those seeking a good introduction into linked data in the cultural heritage domain, but also for the readers who are already familiar with this area. The contributions provide many interesting detail of linked data projects in process and raise questions that will profit from further experimentation.
Melanie Wacker
Metadata Coordinator
Columbia University Libraries
New York, N.Y.