A DOI is a persistent identifier for a document or dataset, standardised as ISO 26324 and operated through registration agencies — Crossref for scholarly literature, DataCite for research data. The identifier resolves to a current location, so a citation survives the publisher reorganising its website.
DOI
The DOI is the most successful persistent-identifier scheme in existence, and its success is organisational rather than technical. The syntax is trivial. What makes it work is that registration agencies operate resolution and metadata as ongoing infrastructure, funded by the members who depend on it.
- Persistence by operation - The identifier survives because somebody runs the resolver, not because the string is clever.
- Registration agencies - Crossref, DataCite and others register, resolve and expose metadata over public APIs.
- Metadata as a public API - Both major agencies publish open, queryable APIs over the full corpus.
- ISO 26324 - Standardised, but the standard is the smallest part of why it works.
The DOI is the anchor of the central argument in The State of Education & EdTech APIs. Education has two standards cultures. One certifies vendors and produces badges; the other runs registries and produces APIs. The scholarly segment built on DOIs, ORCID and OAI-PMH publishes a machine-readable contract 70.8% of the time; the certified classroom software publishes one 15.9% of the time. Operating the infrastructure, rather than certifying implementations of it, is what produced the difference.
Referenced in API Evangelist papers
This standard shows up in my published research. These reports read the machine-readable evidence provider by provider — and put this standard in the context of a real sector.
The State of Education & EdTech APIs
The anchor of the report's central argument: registries that OPERATE infrastructure produced a 70.8% publication rate, against 15.9% where a body certifies vendors instead.
