General track 10: Machine friendly repositories
9:00am - 9:30am
Signposting for Repositories
Los Alamos National Laboratory, United States of America
Repositories holding, for example, scholarly records can easily be navigated by humans. Without much effort, a user can obtain information about the record such as the identifying DOI, the author(s), bibliographic data, and supplemental materials by simply scanning the record’s landing page, for example. For a machine, however, the same task is very hard to accomplish as there are various different conventions as to how to convey this information. The Signposting concept is based on adopting standard web-based mechanisms - typed links and HTTP link headers - to show machines the way to such data. If widely adopted, the repository landscape would become more friendly to machines, closer integrated with the web, and more open to interoperation between systems.
9:30am - 10:00am
Research Graph: Building a Distributed Graph of Scholarly Works using Research Data Switchboard
1Australian National University, Australia; 2National Computational Infrastructure
In this paper, we discuss an open collaborative project called Research Graph derived from the outcome of the Research Data Alliance (RDA) working group on Data Description Registry Interoperability. This project addresses the problem of connecting scholarly works across heterogeneous systems. The RDA working group recommendation provided a solution for connecting publications and research data (data in research) across multiple open access repositories using co-authorship model and jointly funded research projects. Research Graph adopts and extends this work by creating a distributed graph that connects open access repositories to close research management systems traditionally locked behind the firewall. In addition, the distributed graph addresses the challenge of scalability and enables individual universities and repositories to hold a small and manageable graph and synthesis this graph with trusted partner organisations.
10:00am - 10:30am
Trove: Harvesting Australian Repositories
Trove, National Library of Australia
Provides an overview of the Trove service run by the National Library of Australia which harvests collection data from over 400 Australian repositories to provide aggregated access to these collections via Trove. Describes how we use a variety of protocols to harvest and capture data from a diverse range of repositories managed by our content partners across Australia including libraries, museums, archives, universities, historical societies and clubs. Highlights how we provide a number of freely available APIs for use by others and us to be able to generate statistics and re-use collection data and provides some examples of researchers using Trove in support of their research outputs.