Cuellar, Autumn, and Jean Kaplansky. “The State of MathML in K-12 Educational Publishing.” Presented at Balisage: The Markup Conference 2015, Washington, DC, August 11 - 14, 2015. In Proceedings of Balisage: The Markup Conference 2015. Balisage Series on Markup Technologies, vol. 15 (2015). https://doi.org/10.4242/BalisageVol15.Cuellar01.
Balisage: The Markup Conference 2015 August 11 - 14, 2015
Balisage Paper: The State of MathML in K-12 Educational Publishing
Autumn Cuellar has had a long and happy history with mathematics and XML. Her first
degree is in Biomedical Engineering, the obtainment of which involved a love/hate
relationship with Calculus. This degree led to a role as a researcher at the University
of
Auckland in New Zealand. There Autumn co-wrote a metadata specification, explored
the use
of ontologies for advancing biological research, and developed CellML, an XML language
for
describing biological models. Since leaving the academic world, Autumn has been delighted
to share her enthusiasm for XML in technical applications. At Design Science she works
with publishers, engineers, educators, and programmers to implement MathML in XML
publishing workflows.
Jean is an experienced technology trainer, developer, and solutions architect with
a multi-industry background in omni-channel publishing. At Safari, Jean ensures that
published content including books, videos, webinars, journals, and magazine articles
follow content architecture best practices. Jean plays a critical role in the flow
of
content from Safari's publishing partners to Safari's library. Working directly with
publishers, Jean provides expert guidance ensuring the implementation of content
architecture and metadata best practices in content source files. Further, Jean works
with internal Safari teams to continuously optimize tools, systems, and processes;
overseeing new-title content QA. Her publishing production past includes work as the
Senior Director of Solution Architecture for Aptara publishing production, an XML
Content Architect for Cengage Learning, and an XML Consultant at Arbortext. Jean is
a
member of the International Digital Publishing Forum Indexing and EPUB working
groups, as well as the Book Industry Study Group (BISG) Content Structure Committee.
Jean lives out in the woods, just south of the Adirondacks, with her husband and a
menagerie of cats and dogs.
Educational publishers are moving to modern, single-source XML production workflows
and
are embracing MathML as part of the process. The development of single-source publishing
workflows is part of a much larger digital evolution aimed at enabling publishers
to make
content available both as traditional print products and digital content as ebooks.
Furthermore, K-12 publishers are becoming more competitive by including supplemental
materials available through digital-only learning management systems that implement
a
combination of ebook content, teacher lecture material, assignments, assessments,
and
automatically graded homework assignments.
However, educational publishers face challenges in producing math content through
modern
single-source multiple output publishing workflows including:
Support for constructs unique to K-12, such as geometry and visual, mathematical
representations (e.g., diagrams, charts).
Government regulated accessibility requirements and lack of multiple assistive
technology tool support for students (e.g., support for print content disabilities
that are not necessarily a matter of physical disability).
Inconsistent or non-existent MathML support across the major web browser rendering
engines that are the basis of popular desktop and mobile web browsers.
Competing educational publisher business requirements and management
priorities.
Assessment and adaptive learning limitations due to inconsistent or non-existent
MathML support within browsers.
The purpose of this article is to review various publishing challenges specific to
K-12
educational content publishers. A further goal of the article is to provide insight
into
cooperative solutions mitigating current K-12 MathML publishing challenges. The assertions
and recommendations in this article require the fully engaged participation of K-12
educational publishing production management and staff, XML-based markup subject matter
experts, and the web browser rendering engine development community.
Education content publishers increasingly distribute content to multiple delivery
channels including printed textbooks, ebooks, Learning Management Systems (LMS's),
and
web-based Personal Learning Environments (PLE's). To better support new content multiple
destination distribution requirements, education content publishers are increasingly
moving
to XML production workflows and implementing MathML as part of a larger digital evolution.
Publishers traditionally competed by supplementing education programs and content
with
ancillary materials delivered as hard-copy, or increasingly, to an LMS as ebook content,
or
to a PLE capable of automatically grading homework assignments and quizzes. XML is
a
natural fit for storing large volumes of education content including MathML. However,
education content publishers currently face a number of challenges and obstacles to
author,
edit, and distribute math content across delivery channels. This paper describes current
trends in K-12 education content publishing and where MathML fits in through examples
of
how publishers currently address authoring, editorial, and production workflow
obstacles.
Digital Content Revolution
Education publishers and libraries started down the path to digital content revolution
as far back as the mid-1950s with the introduction of the Xerox machine. In 1956,
the
United States Congress approved the Library Services Act (later to be renamed the
Library
Services and Technology Act). This legislation provided public funding to libraries
for
exploring further use of computer-based systems to control bibliographic metadata
and
library catalogs. By 1960, the digital revolution to catalog and store full-text published
documents was in progress, marking the beginning of the end for the Dewey Decimal
System.
The digital content revolution continued as libraries and educators swiftly adopted
computers and databases to manage bibliographic metadata, library catalogs, and the
distribution of scholarly periodical content. Library science in the last twenty years
has
become increasingly about what we can search for and access through a web browser.
Benefits
of the digital content revolution affect libraries and scholarly and education content
publishers.^{[1]}
In 2015, education content publishing, specifically K-12 content publishing, is at
a
turning point. More and more K-12 content is published and distributed as ebooks and
in
online environments as components of LMS's and PLE's than ever before. K-12 publishing
is
now in the very heart of the digital content revolution via the mass adoption of computers,
the internet, web browsers, and ubiquitous devices capable of downloading and rendering
content for student consumption. Distributing content, either scholarly content or
K-12
education content, across the web, has become quite the problem, however.
XML and MathML
Educational publishers are moving to XML-based publishing production workflows and
embracing MathML as part of the industry effort to make content available for both
traditional print channels and digital delivery including ebooks. Furthermore, publishers
are becoming more competitive by providing ancillary education content that often
include
web-based PLE's and adaptive learning systems.^{[2]}
MathML was the very first XML-based markup specification, released in 1998.^{[3]} Both scholarly and education content publishing communities immediately saw the
value of MathML as the means to, at a minimum, markup math content for all the same
reasons
any other content is now marked up as XML. Reuse, repurposing, accessibility, and
multiple
distribution channels are common goals for implementing XML-based publishing production
workflow systems. XML and MathML are a natural fit as a content source format, but
scholarly and education content publishers face some very specific challenges to publishing
math content marked up as MathML^{[4]} including:
Lack of support for K-12 mathematical content, such as stacked equations, long
division, geometry and visual mathematical representations (e.g., diagrams,
charts, number lines).
Lack of annotation support. K-12 math examples often contain associated
line-by-line annotations intended to walk students through problem-solving.
Government regulated accessibility requirements specific to K-12 content, and
the lack of quality assistive technology tool support.
Inconsistent adoption and support for MathML across platforms in web browsers
and reading devices and apps.
Adaptive technology maturity. Adaptive learning scenarios currently only
support very simple automation such as choosing what content to display based on a
student's answer (e.g., multiple choice).
Current Trends in Education Content Publishing
Most industries have been affected in some part by the digital content revolution
(or
the movement of content and publishing production workflow processes into digital
formats)
currently underway, and possibly none more so than education content publishing. Sales
of
physical textbooks and other learning materials account for 70% of education content
publishers’ revenue. Digital content formats now account for 20% of revenue, and bundles,
which usually include a software or web component in addition to a textbook, make
up 10% of revenue.^{[5]}
Digital content formats have seen a 201% revenue growth since 2009. This upward trend
will likely continue as educators realize the flexibility they have to customize and
configure digital content. Digital content formats not only cover ebooks and audiobooks
but
also what might be considered services, including ancillary material such as question
banks
and videos. With a recent general decline in education textbook revenues due to shrinking
school budgets, the education publishing landscape has become more competitive, and
publishers are stepping up by offering a wider range of options for educators. These
days,
K-12 teachers and curriculum planners can customize student textbooks to match curricula
focus, or quickly modify learning materials to accommodate unique needs of their particular
classrooms and students.
As part of the trend towards personalized learning, technology companies are teaming
up
with education publishers to develop new technology systems driven by adaptive learning.
Adaptive learning is a term used for adapting lesson or assessment content to an individual
student based on observations of how the student responds to what he or she is learning.
For example, if a student grasps a concept quickly, the next concept can be presented
without boring the student. If a student struggles with a concept, time can be spent
reinforcing the specific material. Web-based platform providers often use an
assessment-driven approach as the conditional choices made by adaptive learning-enabled
systems. The software compares responses to a benchmark while students work through
exercises. The software algorithmically adjusts remaining exercises in real time,
reflecting the student's current rate of topic mastery. Education technology companies
are
working to make further advances in this area, and publishers are already adding adaptive
learning platforms to their products. Houghton Mifflin Harcourt’s partnership with
Knewton
is one example of a publisher serving the K-12 market with adaptive learning materials.^{[6]}
MathML Content Production Challenges
Though the file storage format for mathematical content has for the most part been
settled by adopting MathML in publishing workflows, education content publishers face
a
number of obstacles when delivering mathematics as part of their digital offerings.
Mathematical typesetting is a tricky beast - often compared to "dark arts" by publishers
and production experts alike. Unlike images, which are often used to enhance or supplement
the text, mathematics is an integral part of the content and loaded with meaning.
Furthermore, the visual layout of a formula must be precise, as this is part of the
mathematical content's meaning. Finally, mathematical equations are often expressed
with
vertical and horizontal layouts. Taken together rendering formulae is a complex challenge.
Despite the fact that MathML is an eighteen-year-old specification, the level of complexity
required to maintain semantic meaning through both layouts and numeric content has
hindered
education publishers' adoption efforts. But education content publishers are now more
motivated than ever to adopt MathML now that the W3C and IDPF formally included MathML
as
part of the HTML5 and EPUB specifications.
It is in this new digital context that education content publishers are expanding
into
new frontiers with digital content but encounter obstacles in math content publishing
production workflows.
Inconsistent MathML Support in Authoring Tools, Web Browsers, and Digital Reading
Devices and Apps
Despite the fact that MathML is part of both the HTML5 and EPUB 3 specifications,
web
browsers and ereading devices and apps are notorious for inconsistent MathML support.
For example, Amazon's Kindle devices and software, the most popular ereading devices
and
software in the United States, do not support MathML. K-12 math content publishers
must
support constructs unique to K-12 education not widely supported in MathML publishing
production workflow tools including authoring and rendering software. Further, K-12
math
content must also meet accessibility regulations, but publishing production staff
often
have trouble convincing management to provide resources required to satisfy regulations
in the face of sometimes conflicting business priorities.
The elevation to recommendation status of the IDPF EPUB 3 and W3C HTML5
specifications respectively in 2011 and 2014 ushered in an exciting new era of enhanced
content. Both specifications now allow the inclusion of various types of media and
the
potential for creating enhanced and interactive digital content has exploded. Ebooks
and
web pages may include video footage, music clips, or small software applications,
just
to name a few media options. EPUB 3 and HTML5 both may also include MathML, opening
new
horizons for publishing mathematical content. For instance, imagine seeing an equation
in a trigonometry etextbook describing a sine curve, tapping or clicking the equation,
and watching the graph of the equation appear with adjustable parameters.
Education content publishers are poised to take advantage of the new specification
capabilities. Unfortunately, MathML support remains inconsistent across both web
browsers and ereading devices and apps. Most ereading software in today's market relies
heavily on browser technology. MathML support in ereading devices and apps may coincide
with support in browser rendering engines, which is the case with Apple's WebKit-based
Safari web browser, and the Apple iBooks reading app for iOS and OS X. However, the
WebKit browser rendering engine only supports a limited portion of MathML 2.0. Apple's
Safari web browser and iBooks can display MathML to some extent on both iOS and OS
X,
but the current level of support is not acceptable for production publishing K-12
math
content.
In contrast, Mozilla's Gecko rendering engine, which is the basis for the Firefox
web
browser, has the most extensive MathML implementation of any web browser to date.
Google
removed all support for MathML from the proprietary Blink web rendering engine in
2013
after announcing the company's intention to move away from WebKit as the basis for
Google's Chrome browser.^{[7]}^{[8]} Moreover, despite Microsoft’s recent stance toward embracing open standards
and the addition of limited support for MathML in Microsoft Office, Internet Explorer
does not support MathML. In fact, Microsoft currently has no plans to implement MathML
in Edge, their latest web browser and Internet Explorer's successor.^{[9]}
Peter Krautzberger (Manager, MathJax Consortium) suggested in an O'Reilly Radar blog
post that the volunteer nature of web browser rendering engine development is a reason
the rendering engine development community has not adopted MathML more widely.^{[10]} Krautzberger notes that Mozilla and WebKit both relied upon volunteer
developers to implement and test MathML implementations. However, volunteers are often
tasked with multiple projects in addition to developing commercial software. It is
not
practical for math content publishers to continue to rely on volunteer effort.
Volunteers may come and go, leaving no one accountable for the results left behind,
and
possibly no one to implement requirements that didn't make it into the web browser
engines in the first place.
The current state of MathML support in web browsers does not preclude future
progress. Mozilla currently supports significant portions of MathML 2.0 and 3.0. Webkit
is close to supporting 80% of the specifications. Conspicuously missing from both
engines, however, is support for elementary MathML constructs: long division, stacked
equations, and K-12 specific visual features (e.g., number lines and various
graphs).
As evidenced by the growing support of MathML in software adhering to the EPUB 3^{[11]} and HTML 5^{[12]} specifications, MathML implementation and support will eventually be
available across web browsers, ereading devices, and apps. All of the major browser
rendering engine vendors and many ereading device and app developers directly
contributed to both the development of the EPUB 3 and HTML5 specifications. The IDPF
and
W3C officially approved including MathML in the recommended specifications. However,
until browsers and ereaders can render MathML natively, MathJax, an open source
cross-platform Javascript library for rendering MathML, is available as an interim
solution.
MathJax is a powerful JavaScript library capable of rendering complex LaTeX as
images, or where possible, as MathML. However, MathJax must modify a web page's DOM.
In
doing so, MathJax rendering can potentially interfere with other libraries and tools
that access the DOM. For example, a braille display that works with iOS may no longer
be
able to decipher MathML if MathJax settings render math as images due to configuration
or browser capability. MathJax fills the gap created by absent MathML support in web
browser rendering engines. MathJax also enables scholars and publishers to create
math
content in LaTeX or MathML. The library has many configuration options and has been
adopted by scholarly societies, and ereading application developers alike. However,
MathJax has a learning curve that developers must address when integrating the library
into applications and platforms. Developers often encounter performance issues that
must
be worked around through specific configurations. Furthermore, the range of possible
MathJax configurations combined with a market where developers no longer know what
screen or device will be used to read content creates a QA risk requiring further
investment to mitigate. Therefore, while MathJax provides a mature solution for
publishing math on the web and in digital content, development and configuration
required to integrate MathJax requires extra "care and feeding" to maintain application
integrations. In the end, the best solution for rendering and displaying math on the
web
and in digital content is still native MathML support in browsers and ereading devices
and apps. Native MathML in web browsers requires maintenance and development on the
part
of browser rendering engine developers, instead of with the scholarly and education
community, who would much rather be writing math than web browser application
code.
Math Requirements Specific to K-12 Education Content
Educators explain sometimes difficult concepts to young minds, and this task is often
best accomplished visually. K-12 math textbooks include a large number of diagrams,
graphs, fill-in-the-blanks, and stacked expressions to teach early math concepts.
See
Figure 1 for an example showing a stacked subtraction
operation with carry-overs, borrowing, and foreground colors to explain large
subtraction operations. Presentation-based MathML markup for all figures in this section
is included in Appendix A.
MathML 3.0 added new features to support elementary math notation, which gets
educational publishers part of the way towards a solution for these more complex
structures. The problem is that since elementary math notation is not widely used
by
scholarly authors and in higher education, tools have been slow to add support for
it.
To address problems with consistency in tool support of MathML, content creators often
have to resort to using the MathML 2.0 approach for encoding equations. For the stacked
subtraction example above, this means using tables and other tweaks. However, the
MathML
specification working group had good reason to add elementary math notation to MathML
3.0 - without it, authors have to make too many manual adjustments to the MathML to
get
it to display properly. Furthermore, MathML 2.0 lacks the semantic information required
by assistive technology to properly render elementary math for students with special
needs.
Once again, we have a situation where tool development must catch up with
specification capabilities. However, even if there were complete web browser
implementations of the MathML 3.0 specification, publishers still have a problem with
heavily image based teaching approaches specific to K-12 math. There is no MathML
markup, for example, to display a number line demonstrating how to add or subtract
with
negative numbers. Therefore, some subject matter experts argue that the highly visual
elements of K-12 textbooks still beyond the scope of the current MathML specification
were never considered in the first place. The general conclusion of these subject
matter
experts is that no implementation should happen before the topic can be revisited
and
prioritized by the MathML working group for inclusion in a future version of the MathML
specification. We do not think this is a practical approach to the implementation
problem and the fact that current solutions disenfranchise thousands of kids from
accessing existing math content due to digital content accessibility support issues.
Authoring and rendering solutions that can support unique K-12 math constructs can
be a
combination of W3C standards: MathML, SVG, and Annotations exist right now.
A combination of MathML and the upcoming W3C Web Annotations specification may be
suitable for some K-12 math only requiring basic rendering and annotations. However,
for
purposes of interactive content delivery, publishers must still rely on rendering
math
as SVG images. Publishers who wish to publish K-12 math content heavily designed and
requiring vector image implementations not currently supported by the existing MathML
specification may require a combination rendering implementation of MathML, SVG, and
Web
Annotations standards together.
The K-12 education publishing community will finally realize a pie-in-the-sky
solution when MathML, SVG, and W3C Annotations are mature enough and combined to create
labeled graphs and equations while retaining the underlying overall advantages of
MathML
solutions. Specifically, distributed digital content that is accessible and rendered
properly in native apps, and devices, as well as the open web platform. The requirement
to combine MathML, SVG, and W3C Annotations to create the most inclusive support for
K-12 math is one reason we continue to advocate for native MathML support in web
browsers and ereading devices and apps.
An additional issue to the K-12 education publishing community is the lack of support
for annotating math concepts in worked through problems with MathML markup as
demonstrated in Figure 7.
Accessibility
MathML enables education content publishers to produce content conformant with
governmental accessibility regulations. MathML is the standard of choice for the
accessibility community because of its detailed description of mathematical expressions.
With the level of detail provided by MathML, a mathematical expression can be converted
to braille, narrated, or navigated by assistive technology in a way that students
with
special needs can comprehend.
In 1974, the U.S. Department of Education established the Federal Individuals with
Disabilities Education Act (IDEA) to provide special services enhancing the education
of
students with disabilities. The U.S. Department of Education provides format guidelines
for states and local school districts for creating accessible content. The Department
of
Education recommends that educational publishers provide materials in the NIMAS format,
which is a national accessible K-12 textbook standard. NIMAS, and other accessibility
standards including DAISY, a standard for digital talking books, and PDF/UA (accessible
PDFs), require mathematical content to be coded as MathML.
Although the standards are in place for encoding math content, it is less clear to
education content publishers how to implement accessibility specifications such that
students have the freedom to use their assistive technology tools of choice. For
example, the popular JAWS screen reading software recently added support for MathML,
but
only within Microsoft Internet Explorer. Further, the implementation is MathJax
dependent. However, as previously mentioned, MathJax may interfere with assistive
technology by modifying the underlying web page structure. The number of potential
assistive technology tools such as braille readers and other screen readers is directly
affected if content providers choose to support the MathML-enabled JAWS screen reading
software.
Publishers currently use tools such as MathSpeak from gh, LLC, MathML Cloud from
Benetech, or MathFlow Equation Composer from Design Science to render MathML as
Text-to-Speech. Screen readers that are not MathML "aware" may then read the rendered
Text-to-Speech content. Figure 8 shows the Text-to-Speech
used as alternative text for the quadratic formula. Using a third party tool to render
MathML for screen readers is a solution that is less than ideal. First, it limits
students in the tools that they can use. For example, Text-to-Speech renditions are
phonetic and, therefore, useless to a braille translation system. Secondly, without
the
ability to read MathML directly, screen readers are not truly rendering MathML. Studies
have shown that students with learning disabilities benefit from a different reading
style than blind students. If speech text is hard-coded for the screen reader, no
room
is left for customizing math speech to specific student requirements.
Figure 8: Text to Speech Quadratic Equation
x equals . fraction negative b , plus minus . square root of b
squared , minus . 4 eh c end root , over 2 eh end fraction
Case Study: Accessible MathML and Personal Learning Environments
Our commitment is to making digital solutions that are usable by the widest
range of students, with or without disabilities.
To fulfilling our continued commitment to accessibility, this is the first step
in moving the dial towards better accessibility of math content. This first step
of authoring math content using MathML is the ideal solution.
Education content publisher with approximately 60 math titles available in
proprietary PLE product as ebooks.
All books contain MathML and corresponding rendered images.
Use Case: Presentation MathML and make it
accessible to screen readers within the personal learning environment
portal.
Display Solution: MathJax to display
MathML. Images for really complex math where required.
Accessibility Solution: MathSpeak from gh,
LLC. to generate screen readable alt text for each MathML instance. Bolt screen
readable alt text descriptions back on to MathML and related images in portal.
Current Status: Implementation is in
progress and ongoing. Publisher's MathML is structured correctly in order to
have gh, LLC. generate screen readable alt text descriptions.
Adaptive Learning
One of the appeals of purchasing digital content for school districts is the
opportunity to free up teaching staff time with automation. Publishers often make
an
item bank available from which staff can draw for exams and homework assignments.
Some
e-learning vendors, such as Madeira Station with its Cognero assessment system, can
even
go so far as dynamically generating mathematical expressions and graphs with a variable
substitution scheme.
On the opposite end, educators are also interested in automated grading of homework
assignments, quizzes, and exams. For multiple choice questions, this is easily done,
but
as students enter upper-level mathematics, where students must provide exercise answers
that are an equation, this becomes more difficult. Currently, most assessment systems
are limited to string comparisons, however. The expectation for the future is for
developers to enhance and grow automatic assessment systems alongside adaptive learning
technology adoption.
In 2015, Adobe Flash has been all but replaced by HTML5-based web applications in
Learning Management Systems that support content, assessment, and assessment-based
adaptive learning that must also support accessibility requirements and regulations.
To
this end, K-12 publishers have retooled and re-engineered the publishing production
workflows required to support systems that automatically prompt students to review
material or present additional assessment questions based on previous answers. Much
of
the retooling and changes to workflow are straightforward: Move content publishing
production workflows away from reliance on proprietary file formats and tools. Maintain
source content in open, markup-based formats enabling the reuse of content in print
books and digital, HTML5-based delivery formats including ebooks and web-based LMS's
and
PLE's.
Moving to markup-based formats enables publishers to easily render education content
to teach children how to read and spell. Math, however, despite being a full-fledged
HTML5 citizen, became a sticky issue for publishers to resolve in adaptive learning
scenarios. Making web-based math problems work in an adaptive learning environment
requires software that can intelligently read metadata and content in order to determine
whether the student has correctly solved the problem and what problem to show next.
Enabling adaptive learning systems to determine whether or not a math problem has
been
answered correctly and what content to show next requires problems to be marked up
such
that software can correctly interpret student responses and process metadata to
determine what content to display based on the response. Figure 9 demonstrates a math problem that, if rendered as an
image, is effectively invisible to adaptive learning systems.
Educational publishers have been publishing K-12 math as images with associated
alternative text for years due to the lack of MathML support in publishing systems
and
web browsers. The "image plus alternative text" approach does not meet the requirements
of today's adaptive learning technology which requires both the math problem and
associated metadata to be marked up for machine readability and the logic of adaptive
learning systems.
How the Education Publishing Community Can Help
Despite the problems with moving K-12 math content into the digital realm, we have
reason to be optimistic. The standards are in place, MathML rendering software exists,
is
continually improving, and in many cases developments are already underway to improve
the
landscape for mathematical communications across the board. Education publishers and
subject matter experts can help by advocating for and supporting MathML authoring
tool and
rendering engine development efforts.
The easiest action education content publishers can take is to let browser and ereader
device manufacturers know the importance of MathML to children’s education. When browsers,
apps, and ereaders support MathML natively, a range of opportunities will open to
not just
publishers, educators, and students, but to anyone who needs to convey mathematical
information electronically.To start, both Google and Microsoft have forums on which
education publisher staff and subject matter experts can register to express interest
in
support for MathML in Chrome and Edge, Microsoft's Internet Explorer successor,
respectively.
A deeper dialog needs to happen to convince browser and ereader vendors to invest
in
MathML development. Education publishers and other elearning software providers can
help by
providing incentives and grassroots support to initiate a dialog with web browser
product
management teams. The United States Justice Department has filed lawsuits against
organizations failing to implement accessibility legislation in the United States.
For the
most part, these lawsuits have been limited to universities.^{[14]} Precedence now exists for a more litigious approach to forcing the publishing
industry and other media entities to comply with accessibility legislation. edX, a
provider
of Massive Open Online Courses (MOOCs), was forced to settle a complaint alleging
violations of the Americans with Disabilities Act filed by the United States Justice
Department in April 2015.^{[15]}
Developers should consider contributing support to MathML by using tools to write
and
implement the specification. One of the greatest benefits of supporting open standards
is
that what works in one application will (or should) work in another application. Developers
can actively prevent disenfranchisement of the overall content audience by adding
support
for native MathML to both native- and web-based applications and by integrating open
source
programs and solutions.
Subject matter experts, developers, and publishing companies may also consider
contributing to projects directly supporting MathML development.Open source projects
that
could benefit from both developer and financial contributions to the Gecko and WebKit web browser engines as well as the MathJax JavaScript
library project.
Conclusion
Education publishers are responding to the combined pressures of decreased spending
in
school districts and increased competition by innovating product and service offerings.
In
addition to traditional print textbooks, education publishers provide ebooks and other
digital content. Digital content delivery formats pave the way for new enhanced educational
experiences with video, software, and interactive media. Furthermore, digital content
delivery formats, especially markup-based formats, can be adapted to the needs of
a
particular district or student. In the United States, federal and state law states
require
all content, including K-12 math content to comply with accessibility legislation.
Publishers continue to face progress-blocking roadblocks when it comes to publishing
K-12 math content in both native apps and on the open web:
Although MathML is a natural encoding format for math content in education,
tool support is lagging for K-12 math constructs in general.
Although MathML is part of both the HTML5 and EPUB 3 specifications, browser
and device vendors do not yet consistently support MathML.
K-12 publishers must support math constructs that are unique to lower level
math content and which are not supported by the MathML 3 specification despite the
added support for very basic and minimal elementary math notation.
MathML rendering engines have varying levels of support for MathML 3.0.
Change is in the air though. Publishers and education institutions now face
potential litigation regarding government regulations requiring the production of
accessible content for students with disabilities. MathML as currently specified is
more
than capable of enabling accessible support for most mathematic disciplines. Education
publishers are just beginning to determine the best approach and workflow to implement
math
content such that students with assistive technology can gain equal access to content
that
was previously unavailable to them.
Education publishers of K-12 math content continue to make do with less-than-ideal
solutions despite existing implementation and adoption issues. The polyglot MathJax
JavaScript library is used widely to support rendering and displaying mathematical
content
in both web browsers and apps driven by web browser rendering engines. For elementary
math
notation, some publishing production editorial teams currently use a combination of
MathML
2 with tweaks and hacks specifically designed to render across browsers through MathJax.
Ideally, these production teams will modify their workflows and current hacks when
the
MathML 3-based elementary math components are supported in editing apps and rendering
engines. Publishers are also tackling the onerous and time-consuming task of producing
alternative text for rendition by screen readers to target a subset of assistive
software.
The future looks bright for open web standards-based mathematical communication in
theory. Native app and web-based mathematical communication success are heavily dependent
on the need for dialog. Recommended participants in this required dialog include education
content provider representatives, members of web browser rendering engines' product
management community, and members of the W3C MathML working group. This dialog has
yet to
occur. That said, MathML-based solution advocates and subject matter experts continue
to
advocate for the implementation and support of delivering accessible and complicated
math
content across multiple rendition formats including native apps and the open web.
MathML
specification and accessibility advocates continue to encourage publishers to start
the
necessary dialog with web browser rendering engine product developers. The purpose
of the
dialog is to discuss further the overall publishing use cases for implementing MathML
natively, as part of all web browser rendering engines.
The outcome of this dialog will be a better understanding of the financial and
implementation resource requirements necessary to render all levels of math, including
K-12
math, for anyone who wishes to learn. Supporting MathML across apps and the open web
is not
a matter of missing parts; it is a matter of prioritization. No manager will prioritize
a
potentially expensive development endeavor without understanding exactly where the
functionality must exist within company priorities. Ideally, managers should prioritize
supporting accessible and complicated math content for all users across a variety
of
digital formats. Litigation should be a last resort to force a company's hand to do
the
right thing rather than the source of re-evaluating business goals and management
priorities.
^{[4]} Fans of the TeX programming language like to think that (La)TeX is the solution to
all math rendering problems, regardless of publication format. TeX has its problems,
however, including the lack of a standardized global implementation on the Web.
Rendering engines exist to process some TeX language variants, but few of these
engines publish TeX as content directly rendered by web browsers. TeX is not useful
in the W3C DOM, limiting overall usefulness in digital content distribution. Further,
TeX is not accessible, and also has unique limitations when it comes to rendering
elementary math. Further discussion of (La)TeX as a solution to K-12 educational math
publishing is beyond the scope of this paper.