Seeding Structured Data by Default via Open Source Library Systems

Abstract

Most libraries use the machine-readable cataloguing (MARC) format to encode and exchange metadata about the items they make available to their patrons. Traditional library systems have not published this data on the Semantic Web. However, some agile open source library systems have begun closing this gap by publishing structured data that uses the schema.org vocabulary to describe the bibliographic data, make offers for items available for loan, and link the items to their owning libraries. This article distills the lessons learned from implementing structured data in Evergreen, Koha, and VuFind; highlights emerging design patterns for publishing structured data in other library systems; and traces the influence these implementation experiences have had on the evolution of the schema.org vocabulary. Finally, we discuss the impact that “the power of the default” publishing of structured data could have on discoverability of library offerings on the Semantic Web.