II · 18

Referrer (rfr)

The Referrer identifies the system that generated the ContextObject — the database, search engine, or citation manager that produced the OpenURL.

The Referrer entity identifies the system that produced the ContextObject. In KEV form the prefix is rfr. The Referrer is optional but is strongly recommended — its presence is one of the few features common to almost every well-formed OpenURL in production.

The info:sid/ convention

By long-standing convention, the Referrer is identified with an info:sid/ URI, whose value is a reverse-DNS-style identifier for the source system, optionally suffixed with a colon-delimited sub-identifier:

rfr_id=info:sid/google.com:scholar
rfr_id=info:sid/example.com:database

The info:sid/ namespace is reserved for naming source systems. The form is informal — no central registry of sid values exists — but the convention is universal and recognised by every link resolver.

Why include it

Resolvers use the Referrer for several practical purposes:

  • Logging and analytics. Knowing where requests come from is essential for measuring usage and for diagnosing link-quality problems.
  • Source-specific handling. Some resolvers apply per-source transformations — fixing known data-quality problems in OpenURLs from specific databases.
  • Knowledge-base feedback. Source identification is also what makes initiatives such as NISO IOTA possible: aggregated statistics on OpenURL link quality, broken down by source.

The legacy sid parameter

OpenURL 0.1 used a flat parameter, sid, with the same purpose. Many sources still emit the older form (often in addition to rfr_id) for compatibility with resolvers that have not been updated. Implementations should accept both.

Sources