Digital scholarly publications as APIs: the case of the Structures of Epic Poetry compendium

Slides:

Abstract

Indices (e.g. locorum, nominum, rerum) facilitate the readers' orientation within a printed publication and help them find the information they are looking for. While the transition to digitalpublishing does not render indices less useful, it does require us to find an adequate adaptation of such indices for a digital publishing environment. What should an index of a digital scholarly publication look like? As argued by McGuire (2013), the natural translation of a printed index in a digital environment is an API (Application Programming Interface) — i.e. a machine-friendly interface aimed at enabling the programmatic access to content published on the Web — and the job of “good publishers of the future” is to provide APIs for their publications.

To demostrate the potential of conceiving scholarly publications as APIs we discuss the case of the Structures of Epic Poetry compendium, a publication edited by Simone Finkmann and Christiane Reitz, and to be published by DeGruyter. The compendium consists of three volumes, with the second volume being further divided into two tomes, and a total of 72 chapters and 60 contributing authors. The sheer number of references to primary sources contained in the compendium (ca. 10,000 references in total) raised considerable challenges for the creation of the index locorum. To tackle this problem, a semi-automatic workflow was devised; the CitedLoci pipeline was used to for the automatic indexing of such references, while student assistants verified and corrected the results it produced. This process allowed us to create a database of cited passages that served multiple purposes. Firstly, it was used to generate the indices locorum, one for each volume, to be included in the printed publication. Secondly, it enabled us to create the EpiBau digital companion, a Web application and API that makes the compendium’s contents more readily and easily searchable by its readers, and publishes part of the raw data on which the individual chapters are based. 2 In particular, the API provides access to the companion's contents by categories epic structures, extracted keywords and cited passages.

We believe that providing digital scholarly publications with appropriate APIs has the potential radically to change the future consumption, discovery and retrieval of publication contents. Such APIs can exist either at the level of single publication – such is the case with this compendium – or can be developed for entire portals, publication series, or even publishers offers. Thanks to these APIs, the accessibility and discoverability of relevant publications can be enhanced by implementing services that provide researchers with publication alerts based on the quotation of specific sources or with links to publications on a specific passage.

References

McGuire, H. (2013). A publisher’s job is to provide a good API for books: you can start with your index, The Indexer 31: 36–8. (Available at https://www.ingentaconnect.com/content/index/tiji/2013/00000031/00000001/art00008).

Romanello, M. (2015). From index locorum to citation network: an approach to the automatic extraction of canonical references and its applications to the study of classical texts. Diss. London. DOI: 11858/00-1780-0000-002A-4537-A.