Author disambiguation for bibliometrics

Associating authors with their research outputs remains a major challenge. The two approaches traditionally used are self-reporting and automated author disambiguation. Self-reporting puts a large burden on authors and usually only works with attractive incentives (such as additional funding) in place. Automated disambiguation is still only at best 95% correct, thus needing manual checks and therefore additional effort. Unique author identifiers offer a third approach to author disambiguation. The ORCID author identifier was launched in 2012, and by 2016 2.5 million researchers have signed up for an account. The ORCID DE project started in Germany in May 2016 will lead to further adoption of ORCID in Germany. The ORCID service combines self-reporting elements (authors always have full record of what is in their public ORCID record) with automated workflows, most importantly the auto-update workflow by Crossref and DataCite launched in late 2015 that will automatically push DOI metadata into an ORCID record if the publisher/data center included the ORCID of the author in the DOI metadata. The presentation will give an overview of the current state of author disambiguation using the ORCID service, including non-textual outputs such as research data and software.     «

Associating authors with their research outputs remains a major challenge. The two approaches traditionally used are self-reporting and automated author disambiguation. Self-reporting puts a large burden on authors and usually only works with attractive incentives (such as additional funding) in place. Automated disambiguation is still only at best 95% correct, thus needing manual checks and therefore additional effort. Unique author identifiers offer a third approach to author disambiguation. T...     »