NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Journal Article Tag Suite Conference (JATS-Con) Proceedings 2018 [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2018.
Abstract
The NCBI Bookshelf is an online archive of over 6000 books and documents in life science and healthcare. Full text content is submitted in XML by participating authors, editors, funders, sponsors, and publishers. This paper describes the requirements for participation: What is the scope of the collection, what types of content are eligible, and what are the scientific and editorial quality standards? Special emphasis is placed on the technical evaluation process: How do participants submit XML and what are the minimum requirements? We will talk about details of the XML submission process, about the challenges and obstacles participants may encounter, and about what NCBI is doing to help during the process.
What is Bookshelf?
Bookshelf provides free access to the full text of books and documents in the biomedical and life sciences as well as health care at the National Center for Biotechnology Information (NCBI), National Library of Medicine (NLM). Content in Bookshelf is made available by participating authors, editors, funders, sponsors, and publishers. Through integration with other NCBI databases, such as PubMed, Gene, Genetic Testing Registry, and PubChem, Bookshelf provides reference information for biological data and facilitates its discovery.
Why Bookshelf?
Bookshelf started in 1999 with the third edition of Alberts’ Molecular Biology of the Cell. At that time, key disciplinary textbooks were added to serve as reference information for biological data archived at NCBI. This was then carried out by linking terms in PubMed abstracts to these key texts. Current textbook authors and publishers choose to submit their textbooks to Bookshelf today not only because it facilitates integration with other NCBI resources, but because it makes their textbooks discoverable by new audiences, primarily through Google and PubMed indexing, or because they wish to provide an open education resource (1,2).
Since then Bookshelf has grown to over 6,000 titles beyond classic textbooks. Another key collection in Bookshelf are open monographs. Bookshelf participants like Springer Nature have found that open access monographs are downloaded, cited, and mentioned online more frequently than non-open access ones (3). Some funders (4,5) and university sponsors (6) have seen Bookshelf as an avenue to make monographs or other non-journal content they’ve funded freely available to increase their impact.
Bookshelf also has a large collection of systematic reviews, technical reports, and other clinical guidelines, and documents. Many of these are submitted by government agencies or other organizations who come to Bookshelf to help make their content discoverable by key audiences they are targeting, such as clinicians or patient advocacy groups. This content is integrated with other NCBI resources when possible; for example, clinical documents on inherited conditions may be integrated with resources like the Genetic Testing Registry (GTR) (7).
What will Bookshelf take?
To be considered for Bookshelf, a work must fall within the scope of core biomedical subjects as defined in the NLM Collection Development Manual (8). The NLM Collection Development and Acquisitions Section (CDAS) reviews proposed content to ensure that it meets this scope criteria (9). Content must be peer reviewed and provided in English as full text. It must also be accepted for publication by a credible publisher or content provider.
Besides textbooks, monographs, systematic reviews, technical reports and other government publications, Bookshelf will also take biographical works, collected works, handbooks and manuals, dictionaries and encyclopedias, standards and guidelines, statistical works, web content, and other types of grey literature (10).
Bookshelf does not take works consisting entirely of original research articles, theses, and dissertations. Proposed content may be written for researchers, health professionals, educators and students, or consumers.
Bookshelf requires that accepted full text be submitted as XML, preferably in the Book Interchange Tag Suite (BITS), although other structured Document Type Definition (DTDs) are permissible. In exceptional cases, Bookshelf may permit submission of Word files that are styled according to the rules for Bookshelf’s in-house eXtyles template for conversion to BITS XML, or as PDFs for agencies and sponsors with whom NLM has an agreement. These Word and PDF files must be suitably styled so that converted XML output meets Bookshelf’s file specifications (11).
Bookshelf requires full text XML because of its mission to 1) archive this content in an accepted archival format, 2) make the content freely accessible and easily retrievable through searches, and 3) allow important segments of tagged content to be easily integrated with other NCBI resources through text mining and other applications.
Who can apply?
Authors, editors, funders, sponsors, or publishers may apply to have their content reviewed. Any application, however, must be permitted by the publisher or copyright holder of the content. All new publishers and sponsors are reviewed by NLM to ensure they follow current best publishing practices, such as methods of selecting, reviewing, and updating content, and adherence to ethical guidelines.
Additional factors NLM considers include the publisher and/or sponsoring organization’s longevity, business practices, corporate structure, and record of performance regarding such issues as: quality of publications; experience in scholarly publishing; involvement with the scientific community; disclosure of and adherence to print and online publication standards; and promotion of editorial integrity and independence (8).
How does Bookshelf decide what to take?
Content review
To be included in Bookshelf, works must meet Bookshelf’s scientific and technical criteria. Once a proposed work is deemed to be within scope for the NLM collection, its scientific and editorial quality is reviewed by a scientific evaluator in the NLM Technical Services Division (9). If the report or book series has already been approved for Medline indexing by the Literature Selection Technical Review Committee (LSTRC), then it does not need to go through a separate Bookshelf content review (12).
NLM will consider the following during a content review (8,10):
- good editorial quality and elements that contribute to the objectivity, credibility, and scientific quality of the content
- the expertise of the authors or editorial panel in the subject area
- content is supported by a range of cited references
- clinical resources have a review cycle for communicating and maintaining their currency
- content is adequately copyedited and structured, usually in the form of well-ordered chapters, sections, and subsections, clearly demarked figures and figure legends, tables, and boxes
Technical review
Once Bookshelf has completed the content evaluation, the technical evaluation stage begins. A provider will now supply sample data, XML files, images, PDFs, and possibly supplementary data files. Bookshelf inspects the data for compliance with its technical requirements (11). Bookshelf will evaluate the sample for completeness, validity to stipulated DTD, quality of data markup, and quality of images (10).
Bookshelf invests a lot of time in evaluating first-time submissions. Submitting full text books in XML is not an easy undertaking. The goal is to provide detailed feedback to submitters up front, so that future submissions will process more smoothly. The approach is simple: Bookshelf staff attempt to move the data through all the production stages, first by running scripts on the content to confirm it is 1) valid, 2) meets style rules, 3) does not contain any table or file integrity errors. Any errors encountered along the way are noted.
The final stage is human quality assurance of the HTML renditions, usually done by comparing the web versions against the PDF or online version. Focus is particularly on ensuring there are not major metadata, structural, reference, or other content issues.
All feedback is compiled into a detailed report and forwarded to the submitter. In this feedback report, Bookshelf differentiates between required fixes and suggested improvements. Required fixes usually concern errors that prevent the document from processing, such as missing images or XML validation errors. The provider must also fix problems that compromise the integrity of the document, such as corrupted text or illegible images. Suggestions for improvements usually relate to poor tagging or cosmetic issues. If a provider fails to correct the required fixes, then the content will be rejected. If the sample is found satisfactory, then Bookshelf will take the content.
What are Bookshelf’s technical requirements?
Data delivery
Bookshelf will set up a password-protected file transfer protocol (FTP) account to which a provider can deposit files. Basic familiarity with an FTP client, for example FileZilla, is required to successfully deposit files. A content provider may already have an existing mechanism to make data available, for example, an FTP service, content syndication, or a download application program interface (API). Bookshelf strongly prefers that providers deposit the data to the NCBI FTP site. Upon request, however, it will evaluate the existing mechanism and may agree to fetch the data rather than requiring submission to NCBI.
Bookshelf requires all data to be packaged in a compressed archive. Acceptable formats include zip, tar, or gzip. Ideally, a provider will supply a separate archive for each document — for example a separate zip file for each book or for each report. While not required, Bookshelf strongly recommends investing in a meaningful naming scheme for the submission packages. Ideally, the name of the archive could be used as an identifier of the document. For a collection of monographs, the ISBN, for example, would be a fitting name for the package. For a set or series of books, a name could be constructed from the series abbreviation and the volume number. Bookshelf receives a great variety of content types; thus far, it has not been practical to institute concrete rules on file naming. Bookshelf works with individual suppliers to come up with a useful naming scheme.
It is very important that the archive files are structured consistently for a given set, series, or other type of document collection. Bookshelf recommends the creation of respective subdirectories for xml, pdf, image, and supplementary files. According to this recommendation, an archive would look like this:
Box 1
978-1-23-456789-0.zip |-- images | |-- 978-1-23-456789-0-F1.tif | |--
978-1-23-456789-0-F2.tif | `-- 978-1-23-456789-0-F3.tif |-- pdf | `--
978-1-23-456789-0.pdf |-- suppl | `-- 978-1-23-456789-0-S1.csv `-- xml `--
978-1-23-456789-0.xmlIt is acceptable to organize the data differently, provided file naming and archive structure remains consistent between submissions. If possible, Bookshelf recommends that the creation of submission packages be programmed or scripted. Scripting ensures consistency and minimizes human errors and inadvertent variations. Scripting also helps to validate the package against a few hard requirements:
- Files should be named uniquely across a package. It is, for example, an error to provide a file fig1.tif both in an images and in a suppl directory. This safeguards against accidental overwriting or confusion of files when they are unpacked and processed further.
- Bookshelf also insists that filenames not contain any whitespace characters. While it is possible to handle spaces in filenames, they do present a complication, for example in command line interpretation or since they need to be escaped in various contexts.
- If the XML references image files or supplementary data files, then the file base name must match the call-out base name in the XML. Likewise, if a PDF is supplied for the document, then the name of the PDF should match the name of the XML file. Bookshelf does not require unparsed entities, such as image files, to be declared in the XML. It relies solely on name matching to link the XML with external files.
Consistent file naming and packaging is especially important for high volume, high frequency submissions; Bookshelf will automate data processing of such submissions.
Images, PDFs, and supplementary data files
For images, PDFs, and supplementary data files, Bookshelf has one basic requirement: The files need to be of sufficient quality to be considered archival.
For images, Bookshelf requires high resolution, uncompressed versions. The image dimensions should be greater than or equal to the intended display size (13). Illegible text, unrealistic colors, or pixilated images are not acceptable. Bookshelf does not prefer a specific format. If the image quality is sufficient and if the files are cross-platform compatible, it accepts common formats such as TIFF, EPS, PNG, JPEG, or GIF. If a complete book or report is submitted, then Bookshelf requires a cover image or an alternative image representing the document to be included.
When signing an agreement to include content in Bookshelf, a provider can choose whether to make PDF versions of the documents publicly available (10). Bookshelf encourages submission of PDF files, even if they are not to be delivered to the public; they are very helpful when performing XML quality assurance. If the PDFs are to be made available on the Bookshelf website, then Bookshelf requires high quality versions. If available, a provider should submit print-quality PDFs.
For supplementary data files, Bookshelf also prefers high quality versions, ideally in portable format. Bookshelf strongly encourages submission of datasets supporting or enhancing the main narrative. “Supplementary data”, however, does not only refer to supplementary material in the classic sense. It can also relate to content that cannot be adequately expressed via the respective XML schema. Typical examples include forms or flowcharts; these are difficult to describe in XML schemas geared toward books, and would require the integration of a specialized XML vocabulary. Such hard-to-convert objects may also be submitted as “supplementary data files”.
XML coding
Providers must submit the full text of a given document in XML. This is the central technical requirement for participation in Bookshelf. The XML must meet several quality standards to be considered acceptable.
General XML requirements
Bookshelf does not require submissions to adhere to a specific XML schema or DTD. It only requests that the XML adhere to a “a mutually agreed upon” specification. This, of course, means a first basic requirement: a schema must exist. Bookshelf does not accept XML without an accompanying validation mechanism. It converts all XML submissions to adhere to BITS (15). Without a machine-readable specification, it would be difficult to develop converters that are able to handle continuous submissions and future documents. Bookshelf keeps a local copy of the DTD or schema for validation purposes, and requests to be notified of any updates to the provider’s specification. What precisely does “mutually agreed upon” specification mean? The following expands on which characteristics an XML specification or vocabulary must exhibit.
Metadata
Any specification acceptable to Bookshelf must provide for sufficiently granular markup of metadata. One of the first things Bookshelf looks at is whether the DTD provides for the tagging of the general components needed to build a simple citation. For a standard book, this means the following elements must be present:
- Title: The DTD must include the title of the book.
- Authors and/or editors: The DTD should differentiate the roles of authors and editors. It should also indicate whether the contributor is a person or an institution. If the former, then it should mark up surname and given name(s) separately. Contributor tagging may only be omitted if the document is authored and published by the same entity.
- Edition: The DTD must include an edition statement of a book. Edition tagging may be omitted if the book is a first edition.
- Publisher name: The name of the publisher must be tagged separately.
- Place of publication: Like the publisher name, an element for the place of publication is required in all circumstances.
- Date of publication: The DTD must include an element describing the publication date of the book. The markup should distinguish the different date parts: year, month, day, or season.
In addition to these bibliographic elements, Bookshelf places great importance on information related to copyright and terms of use. Any DTD must provide for the tagging of a prose copyright statement. Secondly, it must be possible to include a license statement noting the conditions under which the document may be used. If a standard license is used, for example a Creative Commons license, then the DTD should provide for a way to indicate the universal resource identifier (URI) referencing that license. Permissions tagging matters to Bookshelf, because it is tied to the conditions under which content is made available. When signing an agreement, a provider can opt to make the full text XML, and not just the HTML or PDF renditions, available via the NLM LitArch Open Access Subset (14). To avoid abuse or confusion about the terms of reuse, it is important that any copyright restrictions and terms of use are included in the XML itself.
Structural and textual elements
An “acceptable” DTD must also appropriately describe the main structural units of a book. This means, for example, that it must be possible to divide a classic book into front matter sections, parts, chapters, or back matter sections. Nested structures, such as chapters included under broader sections, must be reflected properly in the XML tagging. It should be possible to further divide the main units into smaller subsections, and to express section hierarchy.
Apart from these organizational units, the DTD should allow for markup of common block-level elements. It should be possible to demarcate titles, paragraphs, lists, figures, tables, footnotes, textboxes, block quotes, and reference citations.
Lastly, the vocabulary must also include tags for the main formatting and inline elements, such as italic, bold, or superscript. It should be possible to tag both external links to other resources as well as internal links; for example, from the body of a chapter to a figure or reference.
Examples of DTDs that include all the above-mentioned features include, not only, the BITS and the NLM Book DTD, but also DocBook, and proprietary DTDs from organizations such as Oxford University Press, Springer, Elsevier, or Wiley.
BITS submissions
Bookshelf prefers submissions in the BITS DTD (15). If a content provider creates XML with the sole goal of submission to Bookshelf, then BITS is the recommended format. To supplement the DTD, Bookshelf provides tagging guidelines that describe its preferred XML tagging style for book submissions (16). Submitters can validate their data with the PubMed Central (PMC) style checker, a tool which reports items in a book XML file that do not comply with the tagging guidelines (17). Many of the rules are not specific to book content and target the lower-level textual elements (18). Examples include:
- Markup cosmetics, for example rules against the use of unnecessary empty elements.
- Restriction lists of allowed attribute values, for example, for external link types, list types, or publication date types.
- Object-specific guidelines, for example, when to tag a figure as floating and when as anchored, or how to properly mark up the components in a citation.
Some of the guidelines, however, specifically address issues that arise when tagging book, or book-like content. For instance, they present options to structure different content types or explain how to mark up excerpts for submissions to PubMed when the content does not include an abstract.
Since BITS is only the preferred format, and is not required, it is also not obligatory to use the PMC style checker and the corresponding tagging guidelines. Bookshelf, however, highly recommends using and consulting the checker; the guidelines can answer many tagging questions, and compliance with the PMC style checker will make the submission process smoother.
Supported BITS structures
One of the most challenging issues in books submissions is the question of overall structure. How to mark up the main divisions of a book using BITS? How to tag the main structural units?
As noted, Bookshelf accepts a wide variety of content types. While the system we developed requires a certain level of data homogeneity – meaning it cannot ingest every kind of BITS tagging - it must also be flexible enough to accommodate different kinds of documents. To balance these needs, Bookshelf supports 3 basic structures:
The first model is based on a “standard” book and works also for documents that strongly resemble books, for example, technical reports, extensive reviews, guidelines, or manuals. These documents are usually long, exceeding 50 pages, and have extensive front matter, including a title page, copyright page, and table of contents. The narrative consists of 3 parts: The front matter for textual front materials such as a dedication, foreword, preface, introduction, or contributor list; the body for the main units of the document, for example, chapters, parts, or sections; and the back matter, containing ancillary information such as glossaries, appendices, or reference lists:
Box 2
<book> <book-meta>...</book-meta> <front-matter>
<preface>...</preface>
<front-matter-part>...</front-matter-part> </front-matter>
<book-body> <book-part>...</book-part>
<book-part>...</book-part> </book-body> <book-back>
<ref-list>...</ref-list> <book-app>...</book-app>
</book-back> </book>The second model applies to documents that look more like a journal article than a book. These documents are much shorter, and do not include any lengthy front matter or a title page; the main narrative usually starts on the first page. Examples include short reports and concise reviews, briefs, or separately published summaries, and regularly updated web content. These article-like documents are to be submitted as a <book-part> in a <book-part-wrapper>, the second top-level BITS element. Here, the <book-meta> element does not describe the document itself, but its source: a set, series, or other type of document collection. All metadata for the document fall under <book-part-meta>:
Box 3
<book-part-wrapper> <book-meta>...</book-meta>
<book-part> <book-part-meta>...</book-part-meta> <body>
<p>...</p> <sec>...</sec> <sec>...</sec>
</body> <back> <app>...</app>
<ref-list>...</ref-list> </back> </book-part>
</book-part-wrapper>The third structure is a combination of the 2 previous models. It accommodates documents that exhibit features of both journal articles and books. These are typically reports, papers, or summaries that are issued by government agencies or other organizations, which are not included in any serial collection. They contain substantial front matter, a title page, and a table of contents. At the same time, they are rather short and the main narrative looks very much like a journal article, often including the typical article sections: introduction, background, methods, etc. Documents of this type are to be delivered as a <book> with metadata and front matter included at the book-level. A singe <book-part>, without separate <book-part-meta>, constitutes the main narrative:
Box 4
<book> <book-meta>...</book-meta> <front-matter>
<preface>...</preface>
<front-matter-part>...</front-matter-part> </front-matter>
<book-body> <book-part> <body> <sec>...</sec>
<sec>...</sec> </body> <back> <app>...</app>
<ref-list>...</ref-list> </back> </book-part>
</book-body> </book>Ideally all documents are submitted using one of these 3 models. If this cannot be accomplished, then it should at least be possible for Bookshelf to convert submissions to match one of the 3 structures.
What happens if Bookshelf takes my content?
Before Bookshelf releases a book on the Bookshelf website, a participation agreement must be signed by NLM and the copyright holder of the book. The participation agreement outlines the responsibilities of the participant and Bookshelf, such as a participant’s responsibility to ensure their content on Bookshelf is current. It also requires the copyright holder to indicate whether they agree to make their content available in the NLM LitArch Open Access Subset (10), which permits automated downloading of content via the NLM LitArch FTP service.
Bookshelf also requires that the content provider review an online preview of their content on Bookshelf and approve its release to the public site. This provides the content provider an opportunity to communicate any display issues or errors in the Bookshelf online version.
Once the participation agreement is complete and the preview has been approved, Bookshelf will release the content to its live site. It will be indexed overnight and retrievable by search the next day. Bookshelf also submits bibliographic records to PubMed once a book has been released, unless it is Medline indexed, in which case, it is the content providers responsibility to submit the citations (12). These records should also be available the next day.
If the accepted content is a series or regularly updated online web resource, then the content provider is expected to submit new and updated content once it is available, by posting it to their FTP account. Bookshelf processes this content through scripts and some human quality assurance as necessary. If an issue with the XML or associated files is detected, Bookshelf will request a resubmission of the content to be posted to the FTP. Failure to meet these expectations may lead to termination of an agreement, although Bookshelf tries to first work with participants to communicate problems and collaborate on a plan to address them in a timely manner.
Bookshelf also provides its participants with usage data for their content in Bookshelf, including page views, user sessions, and referrals, with the understanding that it is a mutual effort between Bookshelf and its participants to meet our common users’ needs.
Acknowledgments
We wish to credit our colleagues at the National Center for Biotechnology Information and the National Library of Medicine for their contributions to the Bookshelf project, especially members of the Bookshelf team, as well as Sergey Krasnov, Jeff Beck, Karanjit Siyan, the PMC team of developers, and the NLM Library Operations team of selectors and catalogers. We thank Kim Pruitt, Acting Chief, Information Engineering Branch, NCBI, and Jim Ostell, Director, NCBI, for their vision and support of the Bookshelf project.
References
- 1.
- March 7 NCBI Minute: Textbooks for free on the NCBI Bookshelf [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); February 28, 2018 [cited 2018 Apr 5]. Available from: https:
//ncbiinsights .ncbi.nlm.nih.gov/2018 /02/28/march-7-ncbi-minute-free-textbooks-bookshelf/ - 2.
- Essentials of Glycobiology, Third Edition & New Glycan Website Now Available at NCBI [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); September 7, 2017 [cited 2018 Apr 5]. Available from: https:
//ncbiinsights .ncbi.nlm.nih.gov/?s =glycobiology&submit=Search. - 3.
- Emery C, Lucraft M, Morka A, Pyne R. The OA effect: How does open access affect the usage of scholarly books? [Internet]. Springer Nature White Paper. November 2017 [cited 2018 Apr 5]. Available from: https:
//resource-cms .springernature.com/springer-cms /rest/v1 /content/15176744/data/v3. - 4.
- National Academy of Sciences (US). The National Academies Collection: Reports funded by National Institutes of Health [Internet]. Washington (DC): National Academies Press (US); 1995- [cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/books/NBK4119/ [PubMed: 25121197] - 5.
- Wellcome Trust–Funded Monographs and Book Chapters [Internet]. London: Wellcome Trust; 2011- [cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/books/NBK327155/ - 6.
- Streptococcus pyogenes “A to Zs” Covered in New Book, Freely Available on NCBI Bookshelf [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); March 3, 2016 [cited 2018 Apr 5]. Available from: https:
//ncbiinsights .ncbi.nlm.nih.gov/?s =ferretti&submit=Search. - 7.
- Dietz H. Marfan Syndrome. 2001 Apr 18 [Updated 2017 Oct 12; cited 2018 Apr 5]. In: Adam MP, Ardinger HH, Pagon RA, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2018. Available from: https://www
.ncbi.nlm .nih.gov/books/NBK1335/ [PMC free article: PMC1116] [PubMed: 20301295] - 8.
- Collection development manual of the National Library of Medicine [Electronic manual]. 4th ed. Bethesda (MD): National Library of Medicine, National Institutes of Health, U.S. Department of Health and Human Services; 2004 [cited 2018 Apr 5]. Available from: http://www
.nlm.nih.gov /tsd/acquisitions/cdm/ - 9.
- Fact Sheet: Technical Services Division [Internet]. Bethesda (MD): National Library of Medicine, National Institutes of Health, U.S. Department of Health and Human Services; January 22, 2014 [cited 2018 Apr 5]. Available from: https://www
.nlm.nih.gov /pubs/factsheets/tsd.html. - 10.
- About Bookshelf [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2010-. Information for Authors and Publishers. [updated 2018 Apr 2; cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/books/about/publishers/ - 11.
- About Bookshelf [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2010-. File Submission Specifications. [updated 2018 Apr 3; cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/books/about/filespec/ - 12.
- Fact Sheet: MEDLINE, PubMed, and PMC (PubMed Central): How are they different? [Internet]. Bethesda (MD): National Library of Medicine, National Institutes of Health, U.S. Department of Health and Human Services; February 2, 2017 [cited 2018 Apr 5]. Available from: https://www
.nlm.nih.gov /pubs/factsheets/dif_med_pub.html. - 13.
- Image Quality Specifications [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); [cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/pmc/pub/filespec-images/ - 14.
- About Bookshelf [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2010-. NLM LitArch Open Access Subset. [cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/books/about/openaccess/ - 15.
- Book Interchange Tag Set: JATS Extension, Version 2.0 [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); [cited 2018 Apr 5]. Available from: https://jats
.nlm.nih .gov/extensions/bits/ - 16.
- NCBI Bookshelf Tagging Guidelines [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); [Updated 2018 Feb 21; cited 2018 Apr 5]. Available from https://www
.ncbi.nlm .nih.gov/pmc/pmcdoc/tagging-guidelines /book/style.html. - 17.
- PMC Style Checker [Internet]. Bethesda (MD): National Library of Medicine (US); 2011. [cited 2018 Apr 5]. Available from: https://www
.ncbi.nlm .nih.gov/pmc/tools/stylechecker/ - 18.
- Elbow A, Krick B, Kelly L. PMC Tagging Guidelines: A case study in normalization. In: Journal Article Tag Suite Conference (JATS-Con) Proceedings 2011 [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2011. Available from: https://www
.ncbi.nlm .nih.gov/books/NBK62090/
- What it takes to get into Bookshelf - Journal Article Tag Suite Conference (JATS...What it takes to get into Bookshelf - Journal Article Tag Suite Conference (JATS-Con) Proceedings 2018
Your browsing activity is empty.
Activity recording is turned off.
See more...