Wikidata: Open Linked Data for Library Publishing

Abstract

Wikidata, a collaboratively edited, open, linked data knowledge base hosted by the Wikimedia foundation, includes a growing collection of open citation data. As of November 2018, more than 20 million publications and 160 million citations have been contributed to Wikidata (http://wikicite.org/statistics.html). Many of these data items have been added by bots that contribute data from open bibliographic databases, including PubMed Central, and from data made available by Crossref and the Initiative for Open Citations (I4OC). Although this approach may be the most efficient way to build a large corpus of open citation data, many scholarly journals will be missed. Journals that cannot meet the requirements of a Crossref contract (for financial or technical reasons) will be invisible in growing open citation network. The journals that are likely to be missed are also those that have not been well-served by for-profit publishers and large university presses–including print journals that flipped to open access and journals in fields that are unfamiliar with or unconvinced of the value of a Crossref DOI (e.g., law reviews and some arts and humanities journals). In this presentation we demonstrate how a library publisher can contribute bibliographic data to Wikidata. By using both manual and batch-processing methods, we contributed complete runs for selected journals hosted on our library’s instance of Open Journal Systems. We share our methods for contributing data for journals that mint DOIs and for journals that do not. We also provide a demonstration of the short-term benefits of building this collection in Wikidata and reflect on the challenges of including Wikidata in a library-publishing program.