XML changes lives of print disabled
DCLnews talks to leading accessibility commentator, George Kerscher of the DAISY Consortium, who says we are on the verge of an XML-driven breakthrough that is set to be a "Gutenberg Revolution" to the print disabled.

UP UNTIL RECENTLY, the "print disabled" had few options for reading books. They could find or hire someone to read books for them. They could scan the text and use software that reads text outloud. Or they could get books on tape or in braille. None of these options are ideal -- as George Kerscher discovered over a decade ago when he lost his sight to retinitis pigmentosa, a condition that starts as tunnel vision until the tunnel closes completely.

Because of the progress of the disease, Kerscher gave up work as a high school literature teacher to study computer science. But few of the specialized books he needed had been taped. Even fewer books were in Braille, which he would need to learn. Hired readers didn't get on well when reading complex computer codes and were never there when he needed them. Scanning (OCR) textbooks was, and still is, full of errors and too time consuming to be a real solution to the student.

But even when computer books were available on tape, the typical recorded text ran 30 hours. The only way to look something up was to fast-forward or rewind through piles of tapes.

A Gutenberg Revolution

A man of great determination and ingenuity, Kerscher set about changing this. He and his colleagues eventually figured out how to digitize audio books so that one book fits on one CD-ROM and navigating the book or finding a quote is as easy as searching for a word on a computer. "This is a Gutenberg revolution," proclaimed Nolan Crabb of the American Council of the Blind, and called digital books "the equalizer" that allowed him to get information as quickly as a sighted person.

Kersher is now International Project Manager of the DAISY Consortium, which was formed in 1996 by talking book libraries to lead the worldwide transition from analogue to digital talking books. (DAISY stands for Digital Accessible Information SYstem).

An interview with George Kerscher follows below.

Exclusive Interview with George Kerscher

DCLnews: What problems do students with disabilities have with textbooks?

George Kerscher: It isn't generally known, but at the beginning of the school year, students do not have accessible versions of textbooks. They may show up a week or a month later, or never. This is true in the K-12 arena. At college level it is even worse. In college the diversity of books used means that fewer and fewer people need the same book. The fact is, less than 5% of the books published ever make their way into an accessible format.

DCLnews: Many states have passed laws to help schools and colleges meet their accessibility responsibilities. Could you tell us more about this?

George Kerscher: It started in 1991 when Texas passed the Braille Bill. It required publishers, whose textbooks were adopted in Texas, to provide files to assist with braille production. This spread to other states and was applied to other formats besides braille.

DCLnews: Who advises state legislatures about accessibility and what file formats do they recommend?

George Kerscher: The states consult local people who work in the trenches making braille and other accessible formats. But they don't necessarily know about technological developments in information science. Therefore the advice the legislatures get varies wildly and is based on software that is being used by local production shops. So you see laws that point to Microsoft Word, WordPerfect, or ASCII. Sometimes HTML or SGML, and more recently XML.

DCLnews: The DAISY Consortium is using XML to create a standard national file format -- could you tell us more about this?

George Kerscher: When work on our DAISY 3 file format fell under the National Information Standards Organization (NISO), we decided to go pure XML. We wanted an XML DTD that could be used as a "conversion" DTD -- one where material could be converted to this DTD regardless of where production started. That way, XML processes could be mapped to this DTD and publishers using Quark or other publishing software could have a simple XML target for output. It was intended to be rich enough to provide the access needed by the disability community and easy enough for publishers and disability organizations.

This XML DTD is called DTBook and is only one part of the full ANSI/NISO Z39.86-2002 specification (commonly called DAISY 3). But it does seem to be the best candidate for the standard national file format needed by publishers and the disabled community. DTBook draws heavily from XHTML and adds specific elements for textbooks, such as footnotes, sidebars, front matter and rear matter.

DCLnews: Could you tell us about RFB&D and how they produced a textbook with DTBook?

George Kerscher: Recording For the Blind & Dyslexic (http://www.rfbd.org) is the nation's education library for persons who are blind or print disabled. RFB&D is a full member of the DAISY Consortium. We wanted to actually test what we all believed. So we contracted with a reputable conversion vendor to convert "The American Nation: Beginnings to 1877" into DTBook XML. This high school history book is graphically rich and posed the kinds of real world problems that we knew had to be addressed. We wanted to demonstrate the following:

1. That costs were not astronomical.
2. That the publisher could get an eBook version out of the conversion process (perhaps creating a financial incentive to go to DTBook).
3. How easy it was to conform to the Open eBook Forum (OeBF) Publication Structure Standard and go to eBook formats such as Microsoft Reader.
4. Plus we wanted to hand the DTD to a braille producer so they could see the benefits. And RFB&D wanted to demonstrate making a full text, full audio multimedia version that conforms to the DAISY standards.

DCLnews: What does all this mean for people with disabilities?

George Kerscher: RFB&D launched its digital service on September 3, 2002. This marks the transition from more than 50 years of analog (with the last 25 years being on 4-track cassette). The powerful navigation of the DAISY format makes the cassette obsolete. And I predict a rapid adoption of this technology. Once we start to use text encoded in XML, we can begin to deliver full text synchronized with full audio multimedia product. This dual reinforcement ... see it and hear it at the same time ... is what we believe will make a real difference in the education of all persons with print disabilities. I believe it will revolutionize education for this disability group. No kidding, we are on the verge of a breakthrough that will change the lives of people with print disabilities.

DCLnews: Clearly, there would be a need for people in the United States to be educated and trained in the use of DTBook and/or other XML DTDs?

George Kerscher: As I said earlier, the local people in states who advise legislatures recommend the technology they are comfortable with. They do not know or understand XML or DTDs or what the benefits can be. There is education and training on the correct use of this technology that needs to be spread throughout the nation. Braille producers, audio book producers, and all the organizations and departments who make information accessible to the students need to know about the XML technology that makes conversion to multiple formats so much easier and less expensive. Conversion vendors and publishers also need to understand the DTD.

Guidelines have been produced that describe the elements in terms that should be very clear in the publishing community. References are made to the Chicago Manual of Style and show those constructs using the XML elements. This document is called the Structure Guidelines and is one of the supporting documents that is already in place.

DCLnews: Exactly how does DTBook allow you to produce accessible versions of books more easily, faster, and cost effectively?

George Kerscher: The simple answer is that the text is the key to access. You cannot produce braille without the text. A braille translation package is used that understands the DTBook tag set. In the same way that visual styles are associated with XML elements, braille formatting rules (styles) are applied to the elements. This allows automatic translation into braille. Of course, the trained braille transcriber needs to go through the braille formatted version to add descriptions of graphics and to address the really nasty formatting issues that the automated process cannot handle properly.

In the case of a Digital Talking Book in DAISY format, the text is imported into recording software and narrators use the PC as a teleprompter and record the text. During this process the text and human recording is synchronized. There is also the opportunity to use screen readers and synthetic speech to make the information accessible.

DCLnews Editorial

Read more accessibility related articles at DCL Library

The DAISY Consortium's DTBook XML DTD is part of the ANSI/NISO Z39.86-2002 standard. Details can be found at
http://www.loc.gov/nls/z3986

The DAISY Consortium produced "Structure Guidelines" to accompany the DTBook DTD. Part 2 describes each element and references the Chicago Manual of Style to explain the element's semantics. The Structure Guidelines are available at the DAISY site under publications:
http://www.daisy.org

Lastly the W3C accessibility guidelines can be found at:
http://www.w3.org/wai