Modelling the Provenance of Linked Data Interlinks for the Library Domain

Abstract

As the Web of Data grows, so does the need to establish the quality and trustworthiness of its contents. Increasing numbers of libraries are publishing their metadata as Linked Data (LD). As these institutions are considered authoritative sources of information, it is likely that library LD will be treated with increased credibility over data published by other sources. However, in order to establish this trust, the provenance of library LD must be provided. In 2018 we conducted a survey which explored the position of Information Professionals (IPs), such as librarians, archivists and cataloguers, with regards to LD. Results indicated that IPs find the process of LD interlinking to be a particularly challenging. In order to publish authoritative interlinks, provenance data for the description and justification of the links is required. As such, the goal of this research is to provide a provenance model for the LD interlinking process that meets the requirements of library metadata standards. Many current LD technologies are not accessible to non-technical experts or attuned to the needs of the library domain. By designing a model specifically for libraries, with input from IPs, we aim to facilitate this domain in the process of creating interlink provenance data.