Assigning Persistent Identifiers to Datasets and Other Scientific Resources: An Early Assessment of Success

Research organizations, data repositories, and universities are now assigning persistent web-accessible identifiers to scientific resources, with most efforts utilizing the Digital Object Identifier (DOI) system. Significant progress has been made in the past few years in the development of recommendations, policies, and procedures for creating and promoting these citable identifiers. Little progress has been made, however, in tracking citations to datasets and other digital resources through the research literature. This presentation will outline efforts within the National Center for Atmospheric Research (NCAR) to evaluate the initial success of assigning persistent identifiers to a variety of resources, including datasets, software packages, and research facilities. We will discuss how identification and referencing patterns for datasets and other resources are affected by the assignment of persistent identifiers.

The presentation will also discuss work to develop and evaluate tools for automating the tracing of research resource identification and referencing in the research literature via persistent citable identifiers. Manual processes can be very effective at tracing references to scientific resources within the scientific literature, but are fraught with problems since they are time consuming, require multiple tools (e.g. Google Scholar, literature databases), and do not scale. The human capital required to do this work is far too scarce to be practical beyond small-scale case studies. Automated processes are thus of considerable importance in enabling these traceability efforts to scale as the numbers of identifiers being created for unique scientific resources continues to grow rapidly.