Native Overview

Native entities are the ones where OpenAlex mints its own IDs — W123 for a work, A456 for an author, S789 for a source — and each ID encodes a judgment call about a genuine real-world boundary dispute:

• Are these two records the same work, or two different ones?
• Is this cluster of papers all by one author, or several people who share a name?
• Is “Springer Nature” the same publisher as “Springer Verlag”?
• Does this affiliation string mean the University of Washington or Washington University?

There’s no external authority we can just look up for these; the answer is OpenAlex’s best inference, and it can be wrong. That’s why native entities are the ones you can correct through curation, and why every native entity page has a How we build it section explaining where the records come from, what we do to disambiguate them, and the known failure modes.

The native entities

• Works — every scholarly document. The core entity; everything else connects to works.
• Authors — the people who create works, disambiguated from raw authorship strings.
• Sources — journals, conferences, repositories, and other venues where works appear.
• Publishers — the organizations behind sources, arranged in a hierarchy.
• Funders — the organizations that fund research.
• Awards — specific grants, linking funders to the works they funded.
• Institutions — universities, companies, hospitals, and other organizations authors are affiliated with.

Compare these with vocabulary entities, where there’s no boundary to adjudicate — just a consistent handle on something that already exists crisply.