KBART (Knowledge Bases And Related Tools)

KBART (Knowledge Bases And Related Tools) is a NISO Recommended Practice that standardizes the title-level holdings metadata content providers send to knowledge-base vendors, so that link resolvers and library discovery layers can tell patrons, accurately and in near real time, exactly which journals, e-books, and other titles a given package actually covers.

What problem KBART solves

A knowledge base is the underlying inventory that powers a library’s link resolver (the “Find It” or “Get It” button next to a citation) and, in turn, its discovery-layer search results. That inventory is only useful if it is accurate: which titles are included in a subscribed package, which ISSN or ISBN identifies each one, what date/volume/issue range is actually available, and whether an embargo delays the most recent content. Before KBART, publishers, aggregators, and agents supplied this holdings information to knowledge-base vendors (link-resolver and discovery-service providers such as library systems vendors) in inconsistent, provider-specific formats, which produced broken or misleading links at the point where a researcher expects one-click access. KBART defines a common, tab-delimited data format and a defined set of metadata fields so that any content provider can supply holdings data any knowledge-base vendor can ingest without a bespoke mapping exercise each time.

What a KBART file actually contains

A KBART title list is a plain tab-delimited file, one row per title (or, for serials, per title plus coverage range), following a fixed column header order defined by the Recommended Practice. Core fields include the publication title, print and online ISSN or ISBN, the coverage start/end date, volume, and issue, embargo information, the title-level URL, first and last author or editor (for monographs), and identifiers for the publisher and the content provider itself. The specification also defines conformance levels, so a provider can indicate how completely its file follows the recommended field set, and companion guidance (KBART Automation, NISO RP-26-2019, published 2019-06-18) covers delivering the same data via API/machine-to-machine methods rather than a manually exchanged file.

How it relates to OpenURL and link resolvers

KBART does not replace OpenURL (the link-syntax standard link resolvers use) or link-resolver software; it feeds them. A link resolver decodes an OpenURL request (typically originating from a citation database or discovery search) and checks it against the library’s knowledge base to decide whether full text is available and where to send the user. The knowledge base is only as reliable as the holdings data loaded into it, and KBART is the recommended practice for supplying that data accurately, consistently, and on a predictable update cycle — which is why NISO frames KBART participation as a shared obligation between content providers (who must publish accurate, current title lists) and knowledge-base vendors (who must ingest and refresh them promptly).

Current version and history

KBART was first published as NISO RP-9-2010 (Phase I). It was substantially revised as NISO RP-9-2014, KBART: Knowledge Bases and Related Tools Recommended Practice (Phase II, published 2014-03-27), which remains the current published edition as of this writing. A Phase III revision, circulated in draft as NISO RP-9-202X, went through public comment from September 25 to November 10, 2025, adding support for additional content types (including audio and video), hybrid open-access and global-content handling, an overhaul of the KBART endorsement process, and provisions for a file manifest; as of mid-2026 that revision remains under Standing Committee review and has not yet been published as a final edition. Anyone citing “KBART” in a conformance statement should specify the edition (RP-9-2014, or the eventual Phase III designation once finalized) rather than the acronym alone, since the underlying field requirements differ between editions.

Worked examples

• A journal publisher preparing its annual title-list update for a consortium of library knowledge-base vendors exports a KBART-conformant tab-delimited file listing every active journal title, its print and online ISSN, the current coverage start date, and any rolling embargo period, and submits it through its normal update channel to each knowledge-base vendor it supplies (e.g., the vendors behind major library link-resolver and discovery platforms).
• An open-access aggregator adds several newly indexed journals mid-year and, following KBART Automation (RP-26-2019) guidance, exposes an API endpoint that knowledge-base vendors can poll for incremental title-list updates rather than waiting for the next full file refresh — reducing the lag between a title becoming available and a library’s link resolver correctly recognizing it.

Counter-example

PIE-J (Presentation & Identification of E-Journals, NISO RP-16-2013) is a related but distinct Recommended Practice: it governs how an online journal platform should display ISSN, title-history, and citation information on the journal’s own web pages so that retitled or migrated journals stay findable and citable. PIE-J is about on-page presentation for human readers and citation tools; KBART is about the underlying machine-readable holdings-metadata exchange that populates a library’s knowledge base. A journal platform can be fully PIE-J-conformant in how it displays its title history and still supply an inaccurate or malformed KBART file (or the reverse), because the two practices address different points in the discovery pipeline.

Related terms

See CASRAI’s entries on the NISO Recommended Practice (RP) tier generally, and on adjacent identifier and metadata infrastructure: DOI, ORCID iD, ROR ID, and JATS XML. For how these identifier and metadata systems interoperate in practice, see the identifier crosswalk guide and the Persistent Identifiers & Research Information Systems cluster hub.