Author data, made better together

The Web of Science Group introduces a collaborative approach to improving researcher identity and disambiguation.

The increasingly global approach to research presents unique challenges. Researchers must establish a research identity for their contribution to their field to be recognized – all while competing for funding and positions and establishing valuable collaborative relationships.  The Web of Science Group has designed the latest Web of Science™ release – a new Author Search, Author Record and curation mechanism – to meet these challenges faced by the research community.

A key challenge in establishing researchers’ author identities is name disambiguation across the journals, indexes, and other tools and systems that represent publications. Names are inherently ambiguous; researchers may have a common name shared by hundreds of other people, or their names may have changed over the course of their career. Also, publisher and indexer policies may impose standards on name presentation that lead to different versions of researchers’ names in the published record.

This issue also presents challenges for research administrators and funders, who need to understand the output and impact of their researchers and research teams in order to measure their investment in research initiatives and programs, remain competitive and meet their strategic goals. Within the extended research, publication and funding ecosystem, solving problems around researcher identity will help to improve research efficiency and provide legitimacy to reward and incentive structures.

“Within the extended research, publication and funding ecosystem, solving problems around researcher identity will help to improve research efficiency and provide legitimacy to reward and incentive structures.”

The process of author disambiguation aims to find all publications belonging to a given author and distinguish them from publications of other authors who share the same name. The Web of Science Group has prioritized challenges in establishing and managing researcher identity with our products and services for many years.

  • In 2007, the Web of Science Group (then Thomson Scientific) established the ResearcherID™ framework in order to let researchers begin to take control of their identity in the Web of Science data.
  • In 2012, we recognized the need for a global effort to address this problem and became a founding member of ORCID.
  • In 2017, we acquired Publons, which has become the world’s most comprehensive researcher profile database, showcasing researchers’ full impact across their career, including peer review, publication and citation records, editorial affiliations and contributions, and more.

Today, the Web of Science data ecosystem is closely aligned with both ORCID (via daily feeds of metadata) and Web of Science ResearcherID (now managed by Publons) in order to improve author discovery and drive author name disambiguation for the millions of researchers indexed in Web of Science Core Collection. However, relying on ORCID and Publons only addresses one avenue of feedback on identity and authorship, one that recognizes authors voluntarily claiming their work.

To supplement this work, we’ve developed powerful machine learning algorithms to disambiguate researcher names when there isn’t an external source of feedback. These two avenues of feedback – human curation and powerful algorithms – have helped to establish Web of Science content as a premier source of data for researchers and institutions to make decisions that impact their careers and determine the course of science.

In the latest Web of Science release, we have taken additional steps to improve the depth of research identity. We have introduced an integrated, collaborative approach to help you discover researchers and provide disambiguation feedback. In doing so, we hope to:

  1. Make it easier to find papers authored by the researchers that matter to you.
  2. Empower Web of Science users to submit feedback to curate an author’s publication record in the Web of Science Core Collection.

The Web of Science Group has introduced new features to support the research community:

  • Author Search (Beta): A newly enhanced Author Search intuitively guides you to ‘Author Records’ that provide a set of publications likely authored by a single person. This new search will help you easily find your researchers and their publications in Web of Science Core Collection – regardless of how common or complex their name, or how their name may be presented in different publications over time.
  • New Author Record (Beta): A new ‘Author Record’ provides an initial snapshot of an author’s publications and citation impact in Web of Science Core Collection, and shows alternative names and organization affiliations so you can see how a researcher is reflected in our data over time. Author Records are based on the intersection of a machine learning algorithm, operating on the rich Web of Science metadata and unique citation network, and human curation via Publons and ORCID. They represent a group of publications that are likely the output of a single person, even when a researcher name is common or presented in different versions in the published record.
  • Author Data is now in your control – Because no algorithm is perfect, your feedback matters. While Author Records provide an initial snapshot of authors’ publications in Web of Science Core Collection, your feedback can help fine-tune author identity. When examining an Author Record, any Web of Science user can submit feedback to help further refine any Author Record created by our algorithm.
  • Researchers can claim their Author Record – Authors can now claim ownership of their Author Records and manage their Web of Science identities via Publons, creating their unique Web of Science ResearcherID. Publons provides a 360-degree view of researchers’ activity – giving them a single profile to manage their identities, and the changes researchers makes to their publication record in Publons will seamlessly update their publication records in Web of Science Core Collection and ORCID.

 

Watch now to see these features in action

 

The Web of Science Group continues its commitment to recognizing authorship and providing accurate author data. For more than 50 years, Web of Science Core Collection has indexed every author from every journal we index, along with every cited reference from every article. The consistent structure of our author data and our pioneering work around author identity have laid a solid foundation for further steps in addressing the problems around author identity – helping us to support easy discovery of an author’s published work and assure confident assessment of researcher output and associated citation impact.  This latest effort aligns with our goal of improving research efficiency to speed the course of scientific discovery and innovation.

We are all trying to get to better data. Let’s get there together.

 

Try for yourself: