The spring-board for developing BITS was the observation that a JATS-based XML book model, or a set of related book models, would be useful to a wide variety of publishers of professional and scholarly books, especially but not exclusively to publishers who are already using one of the NISO JATS journal article models and looking for a compatible model for their books. Just as the wide support of the NISO Journal Article Tag Sets enabled many journal publishers to move to XML, a well-supported XML book model (appropriate to their needs and compatible with current citation management tools) would enable book publishers to move to XML and through XML to electronic publication and archiving.

Why add yet another book model to the world? By developing book model(s) based on the existing NISO JATS Tag Suite, we hope to enable publishers to include their books in the systems that already create, manage, publish, and archive their journal articles and to build on the investment they, as well as their suppliers and vendors, have made in learning and developing around JATS.

The goal for BITS is to make a tag set adequate for supporting interchange, archiving, format-conversion, and publishing for scientific, technical, and medical books. As was true for JATS, the intent of BITS is to support marking up the content of material so that it can be reused, repurposed, and made more discoverable. This purpose implies, as it does in JATS, that the ability to reproduce a particular book format is not a goal.

The BITS Book Interchange DTD is a superset customization of the ANSI/NISO JATS Z39.96-2021 Journal Archiving and Interchange Tag Set, with added material to describe STM books, book components such as chapters, and information concerning the inclusion of books and book components in book series. The Tag Set describes both the metadata and the narrative content of a book, both the metadata and narrative content for book components, and collection-level metadata for book sets and book series, when a book part is associated with one or more such collections.

The BITS book model is intended to support scholarly, reference, higher education, medical, and technical books. Just as the NISO JATS journal article models do not attempt to support magazines or any of a wide variety of other serial publications, the book models are not intended describe trade books, cook books, grade school text books, legal documents, or any of the wide variety of books outside the scientific, technical, and medical realm. Although this work is being supported by the National Library of Medicine, this book model should be usable beyond life sciences publishing, as the NISO JATS journal article models are useful in physics, social sciences, linguistics, and poetry.

An explicit goal was the creation of models that would enable the construction of books comprised of articles. The intent was to enable article bodies to pass nearly unchanged into book parts, with changes to only the outer wrapping element and certain book-specific metadata reflecting the move from an issue of a journal to the chapter of a book.

The BITS Book Interchange Tag Set was constructed using the modules in the NISO JATS and adding additional modules to define components that are specific to books. BITS 2.0 was based on ANSI/NISO JATS Version 1.1 (ANSI/NISO Z39.96-2015). BITS 2.1 is based on ANSI/NISO JATS Version 1.3 (ANSI/NISO Z39.96-2021), which is the latest iteration of the original NLM journal article DTDs. This relationship between JATS and BITS has been quite strictly defined: “The models should be as similar as possible and only as different as necessary”. Therefore if JATS has a named structure that also occurs in books, the JATS name (and, to the extent possible, the JATS content model and attributes) were used for BITS directly or extended.

Concerning the many terms used to name components of books (such as chapter, part, unit, module, lesson, segment, division, and section), BITS is entirely agnostic. BITS divides books into book parts and leaves it up to the publisher to call them chapters or units or anything else.

There are, however, a few named book parts in BITS, largely in the narrative front matter and the back matter of a book. Because BITS, like its JATS parent, does not lead publishers but tries to consolidate current publishing practice, the following publisher-requested named book parts have been built into BITS:

• Dedication
• Foreword
• Preface
• Table of Contents (a structural Table of Contents that can be edited)
• Index (a structural index that can be edited)

The named structures for Tables of Contents and Indexes have many specific, unique included elements. In contrast, the named front matter parts, such as “Preface”, are modeled as generic structures. This means that a publisher can choose to use the named book parts or use the element <book-part> to tag all the parts of the book.

Publishers also give many names to collections of books and/or book components, such as book sets, book series, monograph series, and the like. BITS is entirely agnostic concerning such collective nouns and merely collects the metadata naming such a grouping.

There are two top-level elements in the BITS Book:

• the Book element (<book>), to contain an entire document such as a textbook or a monograph; and
• the Book Part Wrapper element (<book-part-wrapper>), to contain a book part such as a “chapter” or “module” that needs to be handled as a discrete unit.

The BITS Book Tag Set is not based on the NLM Book (Bookshelf) Tag Set that was part of the NLM predecessor to JATS. Instead, the BITS book model is based on the NISO JATS Journal Archiving and Interchange Tag Set (known as “Green” from the colors of the Tag Library pages), with book metadata in BITS replacing the journal and issue metadata from JATS Archiving.

By design, there are some structures which have been modeled as elements in other public book DTDs and schemas that were not named explicitly in BITS, including:

• Book Metadata — The book metadata is held in the element <book-meta>, which is not named “front” as is the corresponding element holding the article metadata for a JATS journal article. Book metadata is unlike that for journal articles, making this one of the real changes from ANSI/NISO JATS.
• CCC Statement — There is no element with this name in BITS. In BITS, this may be tagged as a license paragraph (<license-p>) inside a license statement (<license>), which may include the price tagged as a <price>. The @license-type attribute may be set to “CCC-statement” to identify the information.
• Colophon — In BITS, this may be tagged as a paragraph, as a section within the body of the book, as a book part in the back matter of a book, or as a book part.
• Contributor List — There is no special structure for theses lists; they can be tagged using ordinary structures (lists, definition lists, tables, paragraphs, etc.) within a <front-matter-part>, within the narrative front matter, or within the back matter. Such lists should be written in addition the contributor names (<contrib>) listed in the metadata of a book or book part.
• Frontispiece — No special structure was named in BITS; the material should be tagged using ordinary structures within a <front-matter-part> as part of the narrative front matter. While the <styled-content> element “could” be used to capture the special formatting typical in a Frontispiece, this is discouraged. The typical purpose for tagging a Frontispiece is to make the information content discoverable, not to replicate the look and feel of the document.
• Introduction — There will be no explicitly named “Introduction” element, because this name may be applied at many levels: to a front-matter component, a section of the body, or an entire book part.
• Map Group — This NLM-specific element was used in the previous NLM Book Tag Set; it was not replicated in BITS.