What is JATS?

Journal Article Tag Suite

  • A named collection of XML elements with attributes

  • Intended for the description of journal articles

  • Default relationships between elements are given, but can be over-ridden

  • A tag suite not a tag set

  • Many tag sets (a document type) can be built from JATS

JATS is an ANSI/NISO standard: ANSI/NISO Z39.96-2015 JATS: Journal Article Tag Suite (JATS V1.1)

Scope of JATS

Elements for describing the structure of one journal article

  • Originally STEM journal articles

  • Now covers all types of journals: (sociology, economics, poetry, soft sciences, etc.)

  • Top-level element is a journal article, therefore does not model

    • a complete journal issue

    • article-based websites

    • books, pamphlets, magazines, and similar

What JATS models

  • Research articles

  • Review articles

  • Traditionally non-article material such as

    • editorials, columns, essays, and features

    • book, product, and software reviews

    • letters and errata

    • columns of small news blurbs

  • Deliberately out of scope

    • most format-specific, look and feel, pagination elements

    • journal administrative content (TOC, masthead, Author Guidelines, Policy and Scope statements, editorial or advisory boards, etc.)

    • produce, service, job-search, or classified advertising

    • automated testing (CLE, CME, Q&A)

Purpose of JATS

JATS was written for interchange of articles

  • Each publisher/archive/library uses their own schema (DTD/XSD/RNG)

  • They need to get their articles into the same form of XML:

    • to put information into a single repository

    • to exchange information with each other

    • to sell/display items on the same hosting platform

    • so vendors do not need to learn another unique tag set

    • to share tools and resources

JATS is an old-fashioned document model

  • The structural model of a journal article

  • Best fit for purpose

  • Includes:

    • Basic publication metadata

    • Structure for the narrative text, tables, figures

    • Bibliographic references

    • External links to supplementary material, websites, etc.

JATS based on

  • NLM DTD (written in the 1990s for PubMed Central)

  • Analysis of 40+ publisher-specific DTDs (1 schema) and hundreds of journals to:

    • determine common structures

    • determine common metadata

    • determine how to tag references

    • name the elements

  • Models the 80% the publisher-DTDs had in common, and selected structures from the 20% that were different

JATS philosophy

  • Models current articles/publishing practices

    • doe not lead such practice, but reflects it

    • changes as practices change

  • Tries to preserve current order (reading sequence)

  • Should be easy to encode anyone’s journal articles in JATS

  • Does not typically define or set Best Practices

Descriptive not prescriptive

  • Enabling not enforcing

  • Allows all of:

    • very granular markup, e.g., nearly all semantic items in references

    • very chunky markup, e.g., only face markup inside references

    • no markup at all inside references

    • end notes mixed in with references in a bibliography (or not)

Little required, but much is possible

Many ways to tag the same structure

  • Unique identifiers for authors (ORCID) and institutions possible

  • IDs on virtually everything allowed, not required

  • Detailed metadata (e.g., history) allowed, not required

  • Detailed funding reporting possible, not required

  • Linking terminology to ontologies/taxonomies possible

  • Numeration (e.g., list item or section numbers) present in the XML, or not

  • One-place to add an attribute to every element, e.g., RDFa

Popular vocabularies included (not reinvented)

  • MathML (2.0 or 3.0)

  • XHTML tables

  • OASIS/CALS tables

  • Indexing based on DocBook (only better - recursive)

Structure of a JATS journal article

  • Front Matter (<front>)

    • journal-level metadata (journal title and identifier)

    • article-level metadata (article title, author(s), identifiers like a DOI)

  • Body Matter (<body>) the narrative text of the article, for example:

    • paragraphs

    • sections

    • figures and tables

  • Back Matter (<back>)

    • Appendices

    • Bibliographic reference lists with deeply detailed (but completely optional) metadata concerning each reference

JATS’ current niche

Interchange/archiving is still huge, but now also used for journal publishing

  • Publishers creating content

  • Publishers converting backfiles

  • Public archives such as libraries, PubMed Central

  • Private archives such as ITHAKA/JSTOR

  • Hosting services and web vendors standardize on

JATS depth of use/adoption

  • Most middle-sized and small publishers world-wide publish in JATS

  • All of the huge publishers can make JATS for interchange

  • Archives accept (or require) JATS (PubMed Central, US Library of Congress, British National Library, ITHAKA/JSTOR, Australian National Library, many others)

  • Conversion vendors all know how to handle JATS

  • Tools and products for authoring and conversion to JATS from Microsoft Word

  • Numerous web-hosting and service vendors require or support JATS

JATS is how the world publishes journals

JATS is no longer one of the cool kids; it’s just what you do if you have journal articles.

— Jeff Beck at JATS-Con 2017

JATS is in use in at least 25 countries world-wide (US, UK, Germany, France, Australia, Japan, Russia, Brazil, Egypt, etc.)

Multiple JATS public tag sets

Three journal article tag sets made from JATS

  • Archiving and Interchange (very loose, for libraries and archives)

  • Journal Article Publishing (tighter, made for production)

  • Article Authoring (designed for the article author)

Tag sets available in several options*

  • XHTML tables

  • OASIS tables

  • Both table models

  • Either MathML 2.0 or 3.0

*(That’s 10 tag set combinations!)

Three JATS grammar formats available

  • All JATS tag sets are available in DTD, XSD, & RNG format

  • The information content of each is equivalent

  • Therefore:

    • use the most convenient at the moment

    • switch among them if convenient

    • if a tool you want to use prefers one format, use it when using that tool.

This is unimportant. Do NOT spend time or energy on this.

Current JATS governance

(maintenance and funding)

  • NISO Standing Committee oversees development

  • Funded by NCBI (National Center for Biotechnology Information) of NLM (National Library of Medicine)

  • Supported by NISO

  • Mulberry Technologies acts as Secretariat

  • Requests and suggestions made through a public NISO Comment form

https://groups.niso.org/apps/group_public/add_comment.php?document_id=15932

JATS was designed to be customized

Many JATS users use subsets or supersets of JATS

  • Subset — Delete structures/options they do not use to simplify article creation and processing

  • Supersets — Add information for additional metadata, material for processing/business partners

Few JATS users do their own customization

JATS was designed to be customized

By design: High-level philosophy and design parameters

  • DTD is modular (include only what you need)

  • Simple parameter entity mechanism keeps models in sync

  • Customizations isolated, so can compare/change easily

  • Archiving, Publishing, Authoring (in all 10 versions) maintained this way

    • JATS Compatibility model describes how to design/how not to

    • Tag Library documentation describes physical how to customize

    • Consultants teach courses

How are JATS tag sets used now?

  • Many use a tag set out-of-the-box, unchanged

  • Many customize JATS

  • We expected this and planned for it.

  • What we did not expect: Groups of people coming together to make formal JATS supersets

    • BITS Book

      • Book Interchange Tag Suite

      • for books, book chapters, and book collections

    • NISO STS

      • NISO Standard Tag Suite

      • for national, regional, and international standards

    • TaxPub (an extension of the JATS Journal Archiving for markup of taxonomic treatments)

BITS (Book Interchange Tag Suite)

  • BITS is JATS for books

  • A superset of JATS Archiving (Green)

  • Intended for journal publishers already using JATS

  • Two top-level elements: <book> and <book-part> (a book part is a chapter, module, part, etc.)

BITS grew from demand for JATS-compatible Book Model

JATS users, journal publishers, also publish books

  • Want to use familiar (JATS) tools for books

  • Want to mix books and articles in databases and presentation systems

  • Often use articles as book content (e.g., chapter or section in chapter)

  • Want to manage large books in multiple files

  • Publish collections of books (e.g., series)

BITS was developed to meet that need

JATS Archiving used as a starter

  • NLM-sponsored committee-based process

    • included publishers, archives, libraries worldwide

    • started by building requirements and scope

    • built new book, book-part, and collection metadata models on a base as JATS-like as possible

BITS is a much larger tag set than JATS

  • Narrative content largely untouched

  • Book-specific metadata

  • More flexible than JATS, because more variety in books than articles

  • Tools to accommodate large documents (XInclude)

  • Structural Table of Contents

  • Both structural Index and embedded index terms

  • Questions and answers that can be used to construct quizzes/tests

  • Supports cut & paste from JATS (a JATS <article> can become a BITS <book-part> with a few tweaks)

When to use BITS for your books

(why another book tag set?)

  • When you already publish/host JATS journals

  • To regularize an archive composed of JATS articles and books

  • As a conversion target, just like JATS

    • To exchange documents tagged in proprietary book models

    • To create new books and eBooks

  • When your staff and conversion vendors already know JATS

BITS governance

(maintenance and funding)

  • Maintained by the BITS Working Group

  • Currently funded and developed by NLM and used in the NLM Bookshelf project

  • Mulberry Technologies acts as secretariat

  • Suggestions via JATS and BITS mailing lists

NISO STS

ANSI/NISO Z39.102-2017, STS: Standards Tag Suite (Version 1.0)

  • Metadata and the full content of normative standards documents

  • Intended for standards publishing and interoperability

  • May be used for non-normative materials like guides and handbooks, but not designed for them

  • Licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

  • Two top-level elements:

    • <standard> (for standards documents)

    • <adoption> (for standards documents adopting and embedding other standards documents)

NISO STS was based on ISO STS

ISO STS was based on JATS

ISO gave ISO STS to NISO as a starter

  • NISO Committee-based process

    • included people from many types of standards organizations

    • started by building requirements

    • one requirement was backwards compatibility with ISO STS

  • Built new standards model

    • Made metadata richer, more flexible, optional

    • Kept ISO structure but added parallel new material needed by non-ISO SDOs (Standards Development Organizations)

    • Added structures used by a variety of standards organizations (normative notes and examples)

    • Added book-like structures from BITS (Table of Contents, Index)

    • Added onion-like model for adoption of standards (an Adoption has its own material but also contains another standard — entire)

Four Tag Sets built from the NISO STS Suite

  • Two Interchange Tag Sets

    • Interchange allows only XHTML tables.

    • one for MathML 2.0

    • one for MathML 3.0

  • Two Extended Tag Sets

    • Extended allows OASIS Exchange (CALS) tables in addition to XHTML tables.

    • one for MathML 2.0

    • one for MathML 3.0

NISO STS governance

(maintenance and funding)

  • NISO Standing Committee oversees development

  • Funded jointly by:

    • American Society of Mechanical Engineers (ASME)

    • ASTM International

    • and supported by NISO

  • Requests and suggestions made through a NISO Comment form (See below.)

  • Mulberry Technologies acts as Secretariat

https://groups.niso.org/apps/group_public/add_comment.php?document_id=15932

Using the Tag Suites

The JATS Compatibility Meta-Model

Rules and guidance concerning

  • How to extend JATS in ways that are:

    • predictable (know where to find information)

    • consistent (no semantic surprises)

    • generally non-destructive (purpose not compromised)

  • How to keep JATS-spinoffs compatible

https://groups.niso.org/apps/group_public/download.php/16764/JATS-Compatibility-Model-v0-7.pdf

The JATS Compatibility Meta-Model explains

  • What it means for a tag set to be JATS-compatible

  • How a tag set can be developed or modified to be JATS-compatible

    • principles on which JATS compatibility can be decided

    • some specific criteria for JATS compatibility

  • How to extend the JATS vocabularies without conflicting with current JATS vocabularies

  • Those things that should not change for a model to be consistent with JATS

Examples of the Meta-Model principles

Respect the Semantics of existing elements and attributes

Use a named structure to mean the same thing JATS means by that named structure

Linking Direction

Links in JATS go from the many to the one, so, a reference to a section, table, or figure should be an IDREF pointing to the ID of that section, table, or figure

Subsetting

A proper subset of any content model, attribute list, or list of attribute values is always allowed

Meta-Model can be summed up as

  • Do it the way JATS does, or

  • Use a different name for the structure/attributes you intend to use

JATS is not Open Source

  • Tag Sets, DTDs and component modules, and schemas are in the public domain

  • NISO retains control over changes and additions to JATS Blue, Green, Pumpkin, and STS (and their DTDs/schemas)

  • NLM controls BITS (and its DTD/schemas)

But: Anyone may create a new tag set from JATS or use JATS tag sets without explicit permission (because everyone already has permission).

If you customize JATS, you should

  • Mention the Suite if you used it

  • Never modify a JATS module directly

  • Create new modules for new material, so there is no confusion

  • Use the customization mechanisms

  • Include the comment that the JATS schemas request

    • Created from, and fully compatible with the ANSI/NISO Z39.96 Journal Article Tag Suite (JATS).

    • Based in part on, but not fully compatible with, the ANSI/NISO Z39.96 Journal Article Tag Suite (JATS).

Appendix A. JATS Resources

The JATS has been made into a series of tag sets designed to describe a complete journal article. BITS is built on a on JATS base (particularly low-level text-block structures such as paragraphs, lists, figures, etc.) to model a book, monograph, pamphlet, or similar more extended material. NISO STS is a tag set written to describe standards documents for national, regional, international, and organization-developed standards. JATS, BITS, and NISO STS vocabularies are supported by a wealth of documentation, samples, tutorial material, tools, and more. This appendix, while not comprehensive, describes some of the basic resources available to organizations considering or using JATS, BITS, or NISO STS.

The elephant in the room, that the following pages do not include for fear of incompleteness or bias, are the commercial products built specifically for or easily customizable to handle JATS. There are commercial Word-to-JATS and PDF-to-JATS conversion tools, JATS editors, bibliographic reference checkers, products to import JATS into desk-top publishing platforms, and many many more.

A.1 Web Pages for JATS Tag Sets

The most important resources for learning, implementing, or customizing JATS or any of the three JATS Tag Sets are:

  • JATS Tag Libraries Documentation

  • DTDs and schemas (DTD, XSD, RNG)

  • Sample JATS-tagged documents (non-normative)

These resources may be reached from the three JATS Tag Set pages.

Journal Archiving and Interchange Tag Set page

https://jats.nlm.nih.gov/archiving/

Contains links to the documentation and the schemas (DTD, XSD, RNG) for the Journal Archiving and Interchange Tag Set, which is known as Archiving or Green (from the background color of the Tag Library documentation). Archiving is the most permissive JATS sets, allowing the most structures to be captured without change or rearrangement.

Journal Article Publishing Tag Set Page

https://jats.nlm.nih.gov/publishing/

Contains links to documentation and the schemas (DTD, XSD, RNG) for the Journal Publishing Tag Set, which is known as Publishing or Blue (from the background color of the Tag Library documentation). Publishing is a somewhat more prescriptive tag set than Archiving, intended to help publishers and archives regularize their content and facilitating use and reuse of JATS-tagged articles.

Journal Article Authoring Tag Set Page

https://jats.nlm.nih.gov/articleauthoring/

Contains links to documentation and the schemas (DTD, XSD, RNG) for the Journal Article Authoring Tag Set, which is known as Authoring or Pumpkin (from the background color of the Tag Library documentation). Authoring is intended for author use rather than publisher or archive use. To make authoring easier, this Tag Set provides only a single way to tag most structures.

A.2 JATS ANSI/NISO Standard

https://www.niso.org/standards-committees/jats

(links to JATS NISO Standard and NISO Comment Form)

The JATS Standard (ANSI/NISO Z39.96-2015 JATS: Journal Article Tag Suite (JATS V1.1)) describes all the elements and attributes that make up the Suite and the three base JATS tag sets. The fact of the standard is critically important to JATS adoption and use. However, the Standard is not a document to be read to learn how to use or customize JATS. That information resides in the non-normative Tag Library documentation and the DTDs and Schemas posted at NLM and on Github.

A.3 JATS Tag Libraries (non-normative)

The current Tag Library Documentation (in HTML and linked) for each of the three JATS Tag Sets is available online for reading and reference:

Copies of version 1.1 (the current versions as of this writing) to download may be found:

Each of the JATS Tag Sets is documented with a linked HTML Tag Library. Tag Library Documentation contains element/attribute/parameter entity descriptions and models, usage notes, tagged examples, essays on both common and recommended tagging practices, a comprehensive index, a list of where each element can be used, and visual diagrams showing element structures. The basic structure of a JATS Tag Library is:

  • Introduction describing navigation/linking

  • Element pages (one per element)

  • Attributes pages (one per attribute)

  • Context table (where each element can be used; single alphabetical index)

  • Hierarchy diagrams (tree-like visual representations of the content of many elements)

  • An essay about selecting the appropriate JATS Tag Set

  • Essays on common tagging practice (tips, hints, and examples)

  • Index (with synonyms and use-for terms)

  • Notes for implementors concerning changes, customizations. etc.

A.3.1 Each Element Page

Each element is described by a separate HTML page, where the heading for the page displays the element’s tag name followed by its longer descriptive name. The rest of the element description page discusses aspects of the element and its usage.

Table I

Description Defines the element and may provide information on its usage.
Remarks Additional information about the element, explanations of similar or contrasting elements, or instructions for element usage
Related Elements Information about elements associated with or easily confused with the current element
Attributes An alphabetical list of attributes, for elements that may take attributes
Content Model The element’s declaration in XML syntax, i.e., the content of the element
Expanded Content Model The element’s declaration in XML syntax with all parameter entities expanded
Content Description The content of the element, that is, what is allowed to be inside the element, described in natural language
This element may be contained in: An alphabetical list of all elements which may contain the element under discussion (where an element may be used)
Example One or more excerpts from tagged XML documents, showing use of the current element.

A.3.2 Each Attribute page

Each attribute is described by a separate HTML page, where the heading for the page displays the attribute’s tag name followed by its longer descriptive name. The rest of the attribute description page discusses aspects of the attribute, its usage, and the elements to which it can be attached.

Table II

Description Defines the attribute and its purpose (intended use)
Usage Expected uses for the attribute, such as when or why it might be used or how to determine its value
Remarks Additional information about the attribute, such as similar or contrasting attributes, processing information, or relevant material concerning the base element
Used on Elements This section performs three functions: 1) naming the elements which can take the attribute, 2) describing the exact values and defaults for the attribute when used on specific elements, and 3) providing a list of suggested values for those attributes whose value is any character data (i.e., not a specific list of values).
Multiple Elements If the attribute can be used in more than one way, or with several slightly different meanings, there will be more than one Used on Elements section.
Value and Meaning The possible values of the attribute and description of those values.
Restriction Either with the word Restriction or the words Default value and the actual default value.
Suggested usage Suggested values for the attribute, for those attributes whose value is unconstrained (defined as text: letters, numbers, or special characters)
Example One or more excerpts from a tagged XML documents, showing use of the current attribute

A.4 DTDS and Schemas

FTP: ftp://ftp.ncbi.nih.gov/pub/jats/

The latest schema associated with each of the JATS Tag Sets, as well as previous schema versions, are available by anonymous ftp from NLM. The FTP site has one folder for each Tag Set (Journal Archiving and Interchange, Journal Publishing, and Article Authoring) which contains subfolders corresponding to each Tag Set version release.

A.5 Sample JATS-tagged Documents

Many publishers, archives, and web hosting services provide JATS-tagged XML for their articles. A few of these are listed below.

In the Tag Libraries — The first samples JATS-tagged XML most users see are the fragments of sample documents embedded in the Tag Libraries for each Tag Set. Nearly all element pages contain one or more tagged examples; selected attribute pages also contain sample fragments. In addition, complete tagged samples are linked from JATS Publishing (Blue) Tag Libraries.

PubMedCentral Open Access (https://www.ncbi.nlm.nih.gov/pmc/tools/openftlist/) — PubMed Central makes some of its article archive available in JATS XML.

The JATS4R Github pages (https://github.com/JATS4R/JATS4R-Participant-Hub/tree/master/examples) — The JATS4R pages include samples of JATS tagging that the organization used in developing their Best Practice recommendations.

PLOS (Public Library of Science) Articles (https://www.plos.org/) — PLOS articles are available to download in JATS. On the PLOS site, first find the article (search, select), then, beside Download PDF click the down arrow, select XML.

Elementa Articles (https://www.elementascience.org/) — The journal Elementa makes its articles available to download in JATS. On their site, first find the article (search, select), then click the Download in the horizontal bar and select XML.

A.6 JATS Compatibility Meta-Model Description Draft 0.7

https://groups.niso.org/apps/group_public/download.php/16764/JATS-Compatibility-Model-v0-7.pdf

The JATS Compatibility Meta-Model Description is a thought-piece, written to help developers of new JATS-related XML vocabularies create those vocabularies in ways that usefully extend the reach of the JATS vocabularies without conflicting with current JATS vocabularies. This NISO document provides draft guidelines to enable creators and maintainers of JATS-based document models to know when the extensions they make to JATS models are JATS-compatible and to suggest ways in which they can achieve their modeling goals in a JATS-compatible way.

The JATS Compatibility Meta-Model is intended to describe what it means for a tag set to be JATS-compatible and provides guidance on how a tag set can be developed or modified to be JATS-compatible. It provides both the principles on which JATS compatibility can be decided and some specific criteria that must be met to ensure JATS compatibility. Specifically, the meta-model documents those things that must not change about a model for it to be consistent with the JATS models.

A.7 JATS Ecosystem

The ecosystem that surrounds and supports JATS is highly varied. JATS-Con, a conference dedicated to JATS and its daughter vocabularies provides a location for face to face networking, a concentrated source of JATS scholarship, and a permanent archive of the papers and presentations at the conference. Papers about JATS/BITS/NISO STS make appearances at other markup-related events. JATS has an helpful and beginner-friendly discussion list. JATS has an active user organization (JATS4R) that develops Best Practice recommendations. Many organizations (e.g, PubMed Central, JATS4R, and eLife) have made their Schematron and other QA checking and rules available to the public. There are tools and instructions for depositing JATS-tagged references and abstract with CrossRef. Consultants write JATS supersets and subsets and teach courses on how to customize JATS.

Many tools have been written to use JATS in the processing chain, both commercial and open source or small provider tools. Many of these keep source code and documentation on GitHub. There are tools to make PDF into JATS and JATS into PDF or HTML. There are tools to convert MS Word to JATS. There is a JATS flavor of MarkDown.

While there is no single web location where links to all of these resources can be found, some starting places are mentioned in this section.

A.7.1 JATS Guidelines and QA

A.7.1.1 PubMed Central Tagging Guidelines

https://www.ncbi.nlm.nih.gov/pmc/pmcdoc/tagging-guidelines/article/style.html

PubMed Central is a prominent consumer of JATS documents. Their detailed guidelines are, in effect, a description of a JATS subset and use pattern. The PMC Guidelines are published by PMC for their users but have been adopted, in while or in part, by many other users. Guidelines describing PubMed Central's preferred XML tagging style for journal article submissions in the NLM Journal Publishing DTD or the NISO JATS Journal Publishing DTD. Guidelines now include style rules for data tagged in versions 1.1 and 1.0 of the NISO JATS Journal Publishing DTD, as well as both versions 2.3 and 3.0 of the NLM Journal Publishing DTD (JATS predecessor). Note that the style rules differ among the various versions. This site includes links to tools and resources (such as a style checker, fully-tagged samples, fully-tagged citations, etc.) as well as an email distribution subscription list for updates.

A.7.1.2 JATS4R (JATS for Reuse) Best Practice Recommendations

JATS4R (https://jats4r.org/) is an inclusive group of publishers, vendors, and other interested organisations who use the NISO Journal Article Tag Suite (JATS) XML standard.. The organization is based on the principle that JATS is broad and inclusive, but reuse and interchange would be facilitated if JATS users agreed on one (or a small number) of best practices for tagging. Therefore, JATS4R is a working group (of many active subgroups) devoted to optimizing the reusability of scholarly content by developing best-practice recommendations for tagging content in JATS XML.

JATS4R offers Best Practice tagging recommendations for (at least) the following:

  • connecting authors and affiliations,

  • article publication dates versus historical lifecycle dates

  • citations and data citations

  • clinical trials

  • Conflict-of-Interest statements

  • Data Availability Statements

  • display objects

  • math

  • permissions

JATS4R is an active Google Group (https://groups.google.com/forum/#!forum/jats4r) and maintains JATS4R GitHub pages (https://github.com/JATS4R). The Github pages include both samples of JATS tagging (https://github.com/JATS4R/JATS4R-Participant-Hub/tree/master/examples) and a client-side JATS4R validator using Saxon-CE, that JATS users can use to check their compliance with JATS4R recommendations (https://github.com/JATS4R/validator).

JATS4R runs on volunteer labor, and the group welcomes participation by more JATS users. Their aim is to make adoption of JATS4R tagging recommendations as easy and cost-effective as possible for everyone. To do this, they hope to solicit input from many people in scholarly publishing. That input can be in the form of contributing samples, following and commenting on JATS4R recommendations, joining a subgroup to develop recommendations in an area of interest to you, or simply following JATS4R work and helping to spread the word as they develop new recommendations.

A.7.2 JATS-List Discussion List

http://www.mulberrytech.com/JATS/JATS-List/index.html

JATS-List is an open forum for discussion concerning:

  • the Journal Article Tag Suite itself;

  • JATS applications and tools;

  • JATS customizations; and

  • any and all JATS user questions.

JATS-List is open to everyone: users and developers, experts and novices alike. There is no restriction to what may be posted on JATS-List provided it is directly related to JATS. The topics discussed on JATS-List change as new ideas arise or existing problems are dealt with, but the archive contains all of the ideas and solutions that have been discussed on the list.

A.7.3 Conferences and Conference Proceedings

JATS-Con

Conference: https://jats.nlm.nih.gov/jats-con/

Proceedings: https://www.ncbi.nlm.nih.gov/books/NBK65129/

JATS-Con is the annual conference for all things JATS. Sponsored by NLM (US National Library of Medicine) and typically held on their campus at the National Library of Medicine, JATS-CON provides JATS users a forum to tell their success stories and ask questions of the JATS Standing Committee. Topics have included tools and success stories from JATS implementations, the internationalization of JATS, considerations in adopting JATS, converting Math from Microsoft Word into JATS XML, introductions to BITS and NISO STS, and much more. Papers from previous conferences are available in the JATS-Con Proceedings (Journal Article Tag Suite Conference) 2010-2018.

Other Conferences

JATS has also been discussed at Balisage: The Markup Conference (proceedings for 2008-2018 can be found at https://www.balisage.net/Proceedings/index.html). There have also been papers on JATS at Markup UK (London 2018), XML London (2017 and previous years), and XML Prague

A.7.4 JATS Supporting Documentation Page

https://jats.nlm.nih.gov/

This page hold links to the home pages for each of the three JATS tag sets (Journal Archiving, Journal Publishing, and Authoring). This page also links to the JATS ftp site where DTD, RELAX NG schema, or W3C XML schema versions of the JATS Tag Sets are available for download.

In addition, this page provides links to:

  • Home page for BITS,

  • Home page NISO STS,

  • the JATS Wiki,

  • the JATS-Con conference home page, as well as the

  • the JATS FAQ (actually a FAQ explaining the NLM DTD and JATS).

A.7.5 JATS Wikipedia page

https://en.wikipedia.org/wiki/Journal_Article_Tag_Suite

This current and frequently updated page starts with an historical background/timeline for JATS and its NLM predecessor, information on JATS’ technical scope, and a tagged example.

While not complete, this page contains useful information concerning JATS-related tools for creating, editing, converting, and transforming JATS. The JATS tools range from simple forms to complete conversion automation and from free open source tools to commercial products. Tools lists offer such services as:

  • Conversion into JATS

    • Automating document cleanup editing in Microsoft Word and then export to JATS XML (or into BITS, NISO STS, and many other XML DTDs)

    • Converting OpenOffice (LibreOffice) and Microsoft Word documents to JATS

    • Producing JATS output from a scientific document (requires some human support)

    • Converting Markdown to JATS

    • Converting PDF to JATS

  • Conversion from JATS — JATS as input and as output:

    • HTML for display on the web

    • PDF for printing (some tools use XSL-FO, some by HTML5+CSS)

    • JSON for displaying using HTML and Javascript

    • several ePub formats

A.7.6 JATS on Github

While several Github areas might be viewed as a home for JATS, in fact Github hosts numerous JATS initiatives. The list below is representative rather than inclusive. No guarantee is made for how recent the initiatives or tools are nor for their quality.

A.7.7 Crossref Abstracts

https://support.crossref.org/hc/en-us/articles/213126186-NLM-JATS-abstracts

CrossRef Instructions on how to include abstracts imported from JATS/NLM-formatted XML in Crossref deposits.

Appendix B. BITS Resources (Book Interchange Tag Set: JATS Extension)

BITS is built on a on JATS base (particularly low-level text-block structures such as paragraphs, lists, figures, etc.), but describes a model for a book, monograph, pamphlet, or similar more extended material. The intent of the BITS is to provide a common format in which publishers and archives can exchange book content. The JATS, BITS, and NIS STS vocabularies are supported by a wealth of documentation, samples, tutorial material, tools, and more. This appendix, while not complete, describes some of the basic resources available to organizations considering or using BITS. Most tools and processes that handle JATS can also handle BITS.

B.1 Web Page for BITS Tag Sets

https://jats.nlm.nih.gov/extensions/bits/

The most important resources for learning, implementing, or customizing the Book Interchange Tag Suite (BITS 2.0) are:

  • BITS Tag Libraries Documentation

  • DTDs and schemas (DTD, XSD, RNG)

All versions of BITS are available from this National Library of Medicine site. This page includes links to:

  • the BITS Tag Library Documentation (in HTML and linked),

  • the two BITS Tag Sets in DTD, XSD, RNG format, and

  • the JATS-List discussion list, which is where BITS request and comments are posted.

B.2 BITS Tag Library

https://jats.nlm.nih.gov/extensions/bits/tag-library/2.0/

The BITS Tag Library Documentation contains element/attribute/parameter entity descriptions and models, usage notes, tagged examples, essays on both common and recommended tagging practices, a comprehensive index, a list of where each element can be used, and visual diagrams showing element structures. The basic structure of a JATS Tag Library is:

  • Introduction describing navigation/linking

  • Element pages (one per element)

  • Attributes pages (one per attribute)

  • Context table (where each element can be used; single alphabetical index)

  • Hierarchy diagrams (tree-like visual representations of the content of many elements)

  • An essay about selecting the appropriate JATS Tag Set

  • Essays on common tagging practice (tips, hints, and examples)

  • Index (with synonyms and use-for terms)

  • Implementors notes such as customization

B.3 DTDS and schemas

The BITS models and constraints in this Tag Set are encoded in several schema languages: DTD, RELAX NG (RNG), and W3C XML Schema (XSD). All of the schema files for this Tag Set are available by anonymous FTP: ftp://ftp.ncbi.nih.gov/pub/jats/extensions/bits/2.0/

In addition, the schemas are available at the following URIs:

B.4 Other BITS Resources

Appendix C. NISO STS Resources

NISO STS is a tag set written to describe standards documents for national, regional, international, and organization-developed standards. The JATS, BITS, and NISO STS vocabularies are supported by a wealth of documentation, samples, tutorial material, tools, and more. This appendix, while not complete, describes some of the basic resources available to organizations considering or using NISO STS for standards or migrating from ISO STS to NISO STS.

C.1 ANSI/NISO STS Standard

https://www.niso.org/standards/z39.102-2017

The NISO STS Standard (ANSI/NISO Z39.102-2017 STS (Version 1.0)) describes all the elements and attributes that make up the Suite and the three base JATS tag sets. The fact of the standard is critically important to NISO STS adoption and use. However, NISO STS is a complex application and will be challenging to implement without additional documentation beyond what is in the standard. The NISO STS Standard is not a document to be read to learn how to use or customize NISO STS. That information resides in the non-normative supporting materials such as the Tag Library documentation and the DTDs and Schemas.

C.2 Web Page for NISO Standards Tag Suite Supporting Materials

http://www.niso-sts.org/

The most important resources for learning, implementing, or customizing NISO STS Interchange or Extended Tag Sets are the non-normative documentation:

  • NISO STS Tag Library Documentation

  • DTDs and schemas (DTD, XSD, RNG)

  • Sample NISO-STS-tagged documents

C.3 NISO STS Tag Libraries

http://www.niso-sts.org/TagLibrary/niso-sts-TL-1-0-html/index.html

The Tag Library Documentation (in HTML and linked) for NISO STS is available online for reading and reference. The Tag Library provides:

  • descriptions of each element and attribute (identical to those in the standard),

  • remarks on many of the elements and attributes (intended to help users distinguish between similar structures and use the structures correctly),

  • examples showing most of the NISO STS structures tagged in context,

  • explanatory chapters (describing, for example, how to use the tag library, the top-level document elements named in the standard, hot to select the appropriate NISO STS Tag Set, and NISO STS common tagging practices such as options for encoding Terms and Definitions sections),

  • an index, to help users navigate this complex material

  • visual hierarchy diagrams (which provide graphical illustrations of the structure of the more complex elements in NISO STS),

  • a context table (showing where each element may be used), and

  • examples of a standards tagged using NISO STS

C.4 DTDS and schemas

http://www.niso-sts.org/DTDs-XSDs-RNGs.html

The schemas associated with each of the four NISO STS Tag Sets are available. For user convenience, both of the NISO STS tag sets (Interchange and Extended) are provided in DTD, XSD, and RNG formats. And, in all of these formats, each schema is provided with MathML2 and MathML3 versions. The models represented by the DTD, XSD, and RNG forms are logically identical, and interchange is unaffected by the form chosen. Users are encouraged to use the form of constraint language that works best with their chosen tools.

C.5 Sample NISO-STS-tagged Documents

http://niso-sts.org/Samples.html

NISO STS is an enabling tag suite, not an enforcing one. This means that while it provides the means to tag a lot of detail in the narrative body, the references, and the metadata about the standards embedded in the XML file, virtually none of this tagging is required.

In addition, it is often the case that there may be more than one way to tag some parts of standards documents. Selecting one over the other may be a matter of which is most convenient in the creator's production environment or even which they are in the habit of using.

Therefore the tagged samples provided show how one organization tagged one document. They are intended to be illustrative and helpful, but are certainly not the only possible way to tag any of these documents in NISO STS. Examples given include the following.

  • Example of Adoption — Illustrating the adoption mechanism of NISO STS. The standard tagged was published by ISO as ISO 13849-1, and then adopted by CEN as EN ISO 13849-1 (with the addition of several annexes), and then the CEN adoption was adopted by DIN as DIN EN ISO 13849-1 (with the addition of another annex). This document was originally tagged to ISO STS and that many of the new features provided in NISO STS (with the obvious exception of the adoption tagging) have not been added. This illustrates that documents valid to ISO STS are all also valid to NISO STS.

  • Example of a Minimally-tagged Standard — An example of a minimally tagged NISO STS document. All of the content of RFC 8142 is present, and it has been tagged richly enough to support interchange and simple formatting. This example shows that NISO STS also supports relatively light tagging.

C.6 niso-sts-list Discussion List

http://www.mulberrytech.com/STS/NISO-STS/index.html

An open forum for discussion concerning:

  • the NISO Standards Tag Suite itself;

  • NISO STS applications, implementations, and customizations; and

  • NISO STS user questions.

The niso-sts-list is open to everyone: users and developers, experts and novices alike. There is no restriction to what may be posted on niso-sts-list provided it is directly related to NISO STS.

C.7 STS4i (Standards Tag Suite XML for interoperability)

https://github.com/sts4i

A user and community-driven effort to provide tools to work with NISO STS and document best practice recommendations for creating and processing STS XML documents. Although fairly new, this site includes links to:

  • sts4i-tools (an HTML renderer and Schematron ruleset for NISO STS) and

  • additional NISO-STS tagged sample documents.

Appendix D. The JATS, BITS, and STS Begats

Once upon a time (1985-86) SGML was developed and a model for books and articles was needed. Starting with the very first ever public DTD, many models were developed to fill this void, and all had virtues and followers. The NLM DTD was developed for PubMed Central and was widely adopted. In 2001, a study was funded by Harvard/Mellon/Yale to select the best journal tag set, to see if there could be one journal DTD to rule them all, but none was anointed. So NLM donated the NLM DTD to be the basis of a new journal article tag set; NLM and The Mellon Foundation funded the development of that tag set; NLM, Mulberry Technologies, and Inera did the analysis; and Mulberry wrote the first JATS DTD.

  • The NLM DTD begat JATS

  • JATS begat BITS

  • JATS begat ISO STS

  • ISO STS begat NISO STS

The Origin Story for JATS

The first SGML tag set ever written (the Association of American Publishers DTD) was for journal articles. Over time, the AAP journal article DTD morphed into ISO 12083.

Journal publishers were early adopter of SGML, then XML. Many used AAP/ISO-12083, others hired consultants to write their DTDs or wrote their own. The big players all wrote own tag set (as a DTD). Aggregators, conversion shops, and technical services also wrote tag sets. Libraries, archives, and small players wished for a standard tag set.

There were problems with this proliferation of journal article and journal metadata-only tag sets.

  • The large number of tag sets necessitated numerous mine-to-yours conversions.

    • Conversion vendors had to tool up for 100s of tag sets.

    • Electronic archives struggled and wrote many many conversions.

  • Hosters and aggregators pushed their own versions.

  • Many libraries panicked, how could they pick one?

And there was great frustration world-wide.

History of the NLM Tag Sets

Huge Collections of Journal Articles

  • PubMed Central

    • US Congress declared research paid for with US $$ should be available to all

    • PubMed Central (PMC) was developed

    • NLM adopted an industry DTD -- didn’t quite meet needs

    • Decided to create a model

  • Others Were Contemplating Archives of Journal Articles

    • E-Journal Archival DTD Feasibility Study

    • Inera for the Harvard University E-Journal Archiving Project

    • Conclusion: one model (DTD) for all journal articles possible, but did not exist

  • NLM funded development of the NLM DTD to meet both needs:

    • 1999-2001 — NLM’s PubMed Central wrote/used an archival tag set

    • Used for production of PMC

    • Archiving of electronic journal articles in all fields

NLM DTD Widely Adopted

  • Used to submit articles to PMC

    • Converted to XML in NLM tag set after publication

    • Some adopted for XML-based workflows

  • Other archives and libraries adopted

  • Publisher services, conversion vendors, database spinners became familiar with NLM DTD, then began to prefer, then require it

Birth of JATS

http://www.diglib.org/preserve/hadtdfs.pdf

In 2001, Harvard University Library, under a grant form the Mellon foundation, commissioned the E-Journal Archive DTD Feasibility Study to determine if it was feasible to write one domain-neutral tag set for all archived journal article content. The report concluded it could be done, and set forth some of the parameters for such as tag set.

Following that study, a meeting was held at NLM to determine whether a tag set should be developed, as per the Harvard study. Starting from the NLM DTD, with funding by NLM plus Harvard/Mellon, the charter was to:

  • Write a single tag set, with DTD grammar.

  • Preserve the intellectual content of XML/SGML journal articles (but not the look and feel or brand of a journal).

  • Design the tag set to make it easy for publishers/archives to transform documents from their proprietary XML or SGML to this standardized XML for interchange and archiving.

NLM DTD Became JATS

The new generic NLM-DTD was popular and widely used, but NLM owned the NLM DTD. Many would-be users, particularly from non-US countries, needed/wanted an International Standard.

The decision was made to run the standardization process under NISO, as JATS was under NISO. NLM donated the NLM DTD 3.0 to NISO to become a Standard through the NISO committee process and funded the JATS Secretariat. Over the course of the next few years, the Tag Set was renamed to JATS, improved through extensive committee work, and became a NISO Standard (ANSI/NISO Z39.96-2015 JATS: Journal Article Tag Suite) which has been:

  • very widely adopted for internal use by archives, publishers, libraries, and web-hosting vendors, and

  • practically universally used for interchange of journal articles.

Origin Story for BITS

BITS grew from demand for JATS-compatible Book Model. JATS users, journal publishers, also publish books. Archives and libraries also collect and process books. Since most of these organizations were using JATS for their journal articles, they wanted to:

  • Use familiar (JATS) tools for their books,

  • Mix books and articles together in databases and presentation systems,

  • to use articles as book content (e.g., a chapter or a section within a chapter),

  • Manage large books in multiple files

  • Publish collections of books (e.g., series) and

  • Use the same vendors for their books and journals.

BITS was developed to meet that need. NLM sponsored an initial meeting at the NLM Library with JATS publishers, archives, libraries, and vendors from all over the world. At that meeting, NLM agreed to sponsor a committee-based process and the volunteer committee was chosen. The Working Group started by developing requirements and scope. The JATS Archiving (Green, the loosest model) was chosen as a starter set. The Working Group built new book, book-part, and collection metadata models, using JATS as much as possible and designing new metadata for books. Mulberry Technologies acted as the XML hands of the group and built the first BITS DTD.

Origin Story for NISO STS

ISO STS

International Organization for Standardization (ISO) was in the process of improving their internal standards production processes. They wanted to reduce the cost to produce and publish standards, but they particularly wanted to decrease the time it took to publish standards.

At that time, the ISO standards publishing process was:

  • Based on word-processor files (typically Microsoft Word)

  • Slow, and error filled

  • Publication took months/years after standard was completed

  • Electronic versions (other than PDF) were expensive and prone to errors.

ISO studied many XML-based options considered DocBook, TEI, DITA, and JATS. They performed a TEI-based prototype. Following that prototype, and after more consideration, they selected JATS as the base for their XML. They chose JATS Publishing (Blue) as their base model, and built a superset by adding metadata and text structures specific to national, regional, and international standards. By the end of 2011, ISO had revamped its publishing systems and together with Mulberry Technologies, Inc., developed a derivative of JATS to be used for internal ISO standards publishing: the ISO STS (ISO Standard Tag Set).

In building a superset DTD from JATS, ISO:

  • Replaced journal metadata with standards-descriptions,

  • Added local/regional/national tracking,

  • Kept the JATS narrative body structure intact, with just a few additions,

  • Added standards-specific structures such as Notes, Examples, and the TBX term and definition model, and

  • Did not remove anything from JATS. (No element, attribute, structures in the narrative body were removed.)

The original ISO STS DTD was considered to be an ISO-internal tool. It fully modeled structure in use in ISO processing and did not take other types of standards into consideration. There was no consensus-based process in building ISO STS, ISO with Mulberry Technologies and ISO vendors developed what their internal processes needed. The new ISO DTD provided basic metadata for other users (ISO partners):

  • <reg-meta> regional-body metadata

  • <nat-meta> national-body metadata

ISO also developed web-based viewing and distribution and implemented business rules using Schematron schemas. This DTD has been in full production since, with few changes. The DTD and documentation were made openly available at http://www.iso.org/schema/isosts/ and their open use was encouraged. This DTD and much of the tool structure such as the Schematron rules were made public, but not as a standard.

NISO STS

By 2015, several XML models were in use at different organizations to tag standards and standards-type information, including multiple models based on JATS and a number of other independently developed XML models. This proliferation of non-standardized XML models for Standards made interoperability between organizations difficult, increased development costs, and required custom tools.

It was agreed that it would be beneficial to all stakeholders to move ISO STS toward standardization and create an official relationship with JATS. The two standards would then remain coupled and an XML solution for standards could be more easily adopted. It was considered that this initiative would result in greater interoperability and create a foundation for standards utility development, which in turn would aid stakeholders and end users and improve the overall future of standards publishing. So ISO agreed to donate ISO STS and work with the NISO committees to build a new tag set for standards. The NISO STS work item was approved by the NISO Content and Collections Management Topic Committee on June 26, 2015 and by NISO Voting Members on July 31, 2015. Two NISO committees, a steering committee and a technical committee, with representation from many countries, a number of national and international standards bodies and SDOs (Standards Development Organizations) worked to develop NISO STS, which became a NISO standard in 2017.

NISO STS includes four implementations of the suite, called “Tag Set” ’ The Interchanged Tag Sets (one Tag Set for MathML 2.0 and one for MathML 3.0) and the Extended Tag Sets (one Tag Set for MathML 2.0 and one for MathML 3.0). These tag sets are built from the elements and attributes defined in the Suite and are intended to provide models for standards publishing and interoperability. The Extended Tag Sets differ from the Interchange Tag Sets in allowing OASIS Exchange (CALS) tables in addition to XHTML tables.

NISO STS is a fully backwards-compatible superset of ISO STS 1.1. That means that any document that is valid to ISO STS 1.1 will be valid to NISO STS 1.0, by changing only the document type declaration (or other schema identification mechanism) and the @dtd-version attribute on the top level element (<standard>). (Note: This assumes the use of one of the DTDs, XSDs, or RNGs that use the 2.0 version of MathML, since MathML 2.0 and MathML 3.0 are not fully backward compatible.)

The JATS Family Timeline (as of July 2018)

Table III

2003 NLM DTD made public
2008 Last NLM DTD 3.0
2012 JATS 1.0 released (backward compatible with NLM 3.0)
2012 BITS released
2012 ISO STS 1.0 (current is ISO STS 1.1)
2015 JATS 1.1 (current is ISO STS 1.1)
2016 BITS 2.0 (current)
2017 JATS 1.2d1 (Committee Draft)
2017 NISO STS 1.0 (current)
2018 JATS 1.2d2 (current Committee Draft)

B. Tommie Usdin

Mulberry Technologies, Inc.

B. Tommie Usdin is President of Mulberry Technologies, Inc., a consultancy specializing in XML and SGML. Ms. Usdin has been working with SGML since 1985 and has been a supporter of XML since 1996. She chairs the Balisage Markup conference. Ms. Usdin has developed DTDs, Schemas, and XML/SGML application frameworks for applications in government and industry. Projects include reference materials in medicine, science, engineering, and law; semiconductor documentation; historical and archival materials. Distribution formats have included print books, magazines, and journals, and both web- and media-based electronic publications. She is co-chair of the NISO Z39-96, JATS: Journal Article Tag Suite Working Group and a member of the NISO STS working group. You can read more about her at http://www.mulberrytech.com/people/usdin/index.html

Deborah Aleyne Lapeyre

Mulberry Technologies, Inc.

Deborah Aleyne Lapeyre is a Senior Consultant for Mulberry Technologies, Inc., a consulting firm specializing in helping their clients toward better publishing through XML, XSLT, and Schematron solutions. She works with Tommie Usdin as architects and Secretariat for JATS (ANSI NISO Z39.96-2015 Journal Article Tag Suite) and BITS (Book Interchange Tag Suite) and as Secretariat for NISO STS. She teaches hands-on XML, XSLT, DTD and schema construction, and Schematron courses as well as numerous technical and business-level introductions to XML and JATS/BITS. Debbie has been working with XML and XSLT since their inception and with SGML since 1984 (before SGML was finalized as an ISO standard). In a previous life, she wrote code for systems that put ink on paper and used, taught, and documented a proprietary generic markup system named “SAMANTHA”. Hobbies, besides Balisage: The Markup Conference, include pumpkin carving parties.