VI · 38

Link Resolvers

A link resolver is the institution's gatekeeper for context-sensitive linking — it parses an incoming OpenURL, consults a knowledge base, and routes the user.

What a link resolver is

A link resolver is a server-side service operated by a library, a library consortium, or a vendor on a library's behalf. It receives incoming OpenURL requests, parses the ContextObject, and decides what to deliver. The decision is made against a knowledge base — a per-institution catalogue of titles, subscriptions, holdings, and licensed full-text targets — and a configurable set of policies.

What it does with an OpenURL

For each request, the resolver:

  • Parses the query string, separating administrative parameters from entity descriptors.
  • Identifies the Referent — by identifier when one is present, by descriptive metadata otherwise.
  • Looks up the Referent against the knowledge base, gathering candidate targets: full-text aggregators, publisher sites, open-access repositories, the library's own catalogue, ILL request forms.
  • Ranks and filters the candidates by institutional preference and licence terms.
  • Either redirects the user directly to a preferred target (a "direct link") or presents a menu of options.

Major products

Several products dominate the link-resolver market. Listed here without endorsement; specific feature comparisons are out of scope for this reference.

  • SFX (Ex Libris, part of Clarivate). The original commercial link resolver, descended from the Ghent prototype. Long history of installation in research libraries.
  • 360 Link (Ex Libris, part of Clarivate). Originally Serials Solutions; widely deployed at academic libraries.
  • Alma / Primo (Ex Libris, part of Clarivate). The unified resolver inside the Alma library services platform and Primo discovery service.
  • EDS (EBSCO Discovery Service). Includes its own resolver, EBSCOhost LinkSource.
  • WorldCat Discovery (OCLC). Uses OCLC's WorldCat knowledge base for resolution.
  • Umlaut (open source). Developed and maintained at Johns Hopkins; used at several research libraries.

All these products accept both OpenURL 0.1 and 1.0 syntax. All accept the SAP1-conforming form of 1.0 — KEV-encoded, By-Value, HTTP GET.

Knowledge bases

A resolver is only as accurate as its knowledge base. Vendors operate large, shared knowledge bases (OCLC's WorldCat knowledge base, Ex Libris's Central Discovery Index, EBSCOhost Holdings & Linking Manager) that participating libraries customise to their own holdings. The KBART NISO recommended practice exists in part to improve the quality of data flowing into these knowledge bases. See OpenURL & KBART.

Sources