Crossref has introduced a new method to enhance metadata for scholarly publications by automatically assigning research funder names to standardized identifiers from the Research Organization Registry (ROR). The primary objective is to resolve the problem of inconsistent and often inaccurate text entries for funders, which complicates the analysis of research funding.
Crossref is planning a transition from its own Funder Registry to the open ROR registry, which is utilized across the global research infrastructure. This shift facilitates a better connection between information on institutions, grants, and publications within a unified data ecosystem.
The new system uses a combination of search and heuristics to identify the corresponding organization in the ROR registry from free text. It is designed to be “cautious”: if the system is uncertain, it prefers not to assign any identifier at all. Thanks to this approach, it achieves an exceptionally high precision of approximately 99%, albeit with a lower recall of about 81%.
This project serves as the first practical implementation of a new metadata matching infrastructure that Crossref intends to use for other tasks, including the pairing of affiliations, grants, and citations. The ultimate goal is to provide a richer and more precisely linked overview of scientific research and its funding.