TitlePages - Preamble - Introduction - Accessibility - UniversalAccessibility - PostProduction - Metadata - Accessibility Metadata - PNP - DRD - Matching - Interoperability - Conclusion - References - Appendix 1 - Appendix 2 - Appendix 3 - Appendix 4 - Appendix 5 - Appendix 6 - Appendix 7 - Appendix 8 - Appendix 9 - Appendix 10 - Appendix 11
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 2.5 Australia License.
© 2008 Liddy Nevile
In the previous chapter, there is an introduction to the complex scene in which the research has been grounded. It is never easy to know where to start, in complex circumstances, so some preliminary comments were offered in the hope of painting a broad-brush picture into which details could be added as required.
In this chapter, greater detail is offered with respect to the research and the topics under consideration. As the thesis progresses, definitions are enriched but early on there must be some provision for the reader to gain an understanding of the perspectives of the author.
The first decade of international effort to make the Web accessible has not achieved its goal and a different, or at least complementary approach is needed. The Web has evolved significantly in its first decade as has its use. In order to be more inclusive, the Web needs published resources to be described to enable their tailoring to the needs and preferences of individual users. Also, resources need to be continuously improvable according to a wide range of needs and preferences. Thus, there is a need for management of resources and their components. This research asserts that can be achieved with the correct use of metadata.
The specification of a metadata framework to achieve such a goal is complex. The requirements are not determined simply by the end use, but as an environment in which metadata is specified for an end-use. As the requirements for that end-use cannot be specified in advance, this is a meta-design activity. It is transdisciplinary and here, undertaken in a socio-technical context. The product is better understanding of metadata, a theoretical contribution to the field.
This thesis asserts that the low level of accessibility of the Web justifies a new approach to accessibility and that the most appropriate is a comprehensive process approach that brings together a number of strategies for use according to the circumstances and context. In particular, it should be possible to continuously improve the accessibility of resources and for this to be done by third parties, independently of the original author. This continuous improvement cycle, in turn, depends on the availability of metadata to manage the process. Metadata 's role in management is not new but perhaps is not as well known as its use for discovery.) How that metadata should be developed is at the core of this thesis. The research responds to the need (documented in Chapter 4) for an effective new approach to accessibility.
The general aim of work in the accessibility field is to help make the information era inclusive. Inclusive is a term used in this context to refer to a particular approach to people with disabilities and to the disabilities themselves. People with accessibility needs are not homogenous and many of them do not have long-term disabilities: what they need now may not be of interest to them in different circumstances or at other times. Accessibility is also a special term in this context, designating a relationship between a human (or machine) and an information resource. Both terms are defined in Chapter 3.
The research starts with a close examination and analysis of current accessibility processes and tools. It moves on to include a new approach that will complement previous accessibility work. Then it addresses the problem of how to develop a metadata framework to support a more process-oriented approach to accessibility. Co-editing of international specifications and standards for accessibility metadata, known as AccessForAll (AfA) metadata, was undertaken simultaneously with the research to determine metadata recommendations for several contexts including a Dublin Core Metadata Application Profile module (see Chapter 7).
Actively promoting accessibility is taken to mean being inclusive. The term inclusive is used for operations and organisations that follow appropriate practices to promote accessibility and accommodate improvements in a constantly widening range of contexts. The new process work suggests a 'quality of practice' approach to the process of content and service production that will support incremental but continuous improvement in the accessibility of the Web and thus inclusion in the digital information era.
The thesis chapters report on:
In this section, there are brief introductions to the major terms and concepts used in the research. These are further refined in later chapters.
The United Nations Convention on the Rights of Persons with Disabilities and its Optional Protocol (UN, 2006) calls for equity in access to information and communications. In this thesis, the information and communications of concern are those that are digital and electronic and the terms are used both as nouns and as verbs: people need access to hardware and software to create, store, and deliver digital files as well as to the intellectual content of the files themselves. Collectively, these constitute what is called 'the Web' in this thesis; the Web of digital information and communications.
In particular, the Web is not simply the 'pages' that are encoded in HyperText Markup Language [HTML 4.01]. While such pages might provide the 'glue', it is clear that the information and communication enabled by them is most likely to be made available in a wide range of forms. A typical and simple example of an HTML-encoded page is provided by a temporary 'homepage' of a newly elected Australian Prime Minister (Figure 5).
On this very small Web page (Figure 5), there are six links that put the user in contact with other 'pages' as we might call them. To contact the Prime Minister, one does not send email that would be easily accessible but, instead, receives another page with a form on it. The form saves the Prime Minister from receiving email directly from the user but it also introduces an accessibility issue; many forms within standard HTML pages are not accessible, as defined herein.
Links are provided on the Prime Minister's page to three sources of information that explain privacy, copyright and about the site. One link directs the user to the archive of the previous Prime Minister's Web site. This is a substantial source of information and when contact is made, it reveals files in a range of formats. This archive is provided by the National Library of Australia and before choosing a version, the user can see metadata associated with the archive describing the formats of files involved. (Interestingly (one might say typically), the note does not always display properly, even on a common user agent such as Safari (the native browser for Apple Macintosh computers) (see Figure 6)).
Only when the 'correct' font size is used is the full note legible. The so-called ‘correct’ size is that selected by the resource author, without reference to the user’s needs or preferences.
Figure 7 shows the range of applications necessary to access what is on the first page of the archive but then, each page of that archive is likely to point to yet more resources. All of these resources, the hardware and software needed to use them, form what in the research is defined to be 'the Web'. In fact, the Web might include documents to which there is no Web access, such as paper documents in the National Library, but they would be included as part of the Web because they are linked via their metadata. It can also include people, referrred to by URIs.
In 2004, Tim O'Reilly described the Web using a new term that has since become a model for describing recent versions of evolved products that in fact have no formal versions. Later he said of it (2005):
The concept of "Web 2.0" began with a conference brainstorming session between O'Reilly and MediaLive International. Dale Dougherty, web pioneer and O'Reilly VP, noted that far from having "crashed", the web was more important than ever, with exciting new applications and sites popping up with surprising regularity. What's more, the companies that had survived the collapse seemed to have some things in common. Could it be that the dot-com collapse marked some kind of turning point for the web, such that a call to action such as "Web 2.0" might make sense? We agreed that it did, and so the Web 2.0 Conference was born.
In the year and a half since, the term "Web 2.0" has clearly taken hold, with more than 9.5 million citations in Google. But there's still a huge amount of disagreement about just what Web 2.0 means, with some people decrying it as a meaningless marketing buzzword, and others accepting it as the new conventional wisdom.
The Web has been envisioned by O’Reilly (2005) as a ‘platform’, an integrated entity:
Like many important concepts, Web 2.0 doesn't have a hard boundary, but rather, a gravitational core. You can visualize Web 2.0 as a set of principles and practices that tie together a veritable solar system of sites that demonstrate some or all of those principles, at a varying distance from that core.
O'Reilly (2005) offered the following diagram (Figure 8) to help others visualize this 'new' Web.
Figure 8 shows many interactive 'spaces' (grey) as part of the Web. This means that users do not just receive information and communications but they initiate or respond to them as well. For this, they need a range of competencies (orange). The Web, as it is now, has a number of features (pink).
Web 2.0, the current Web, is vastly different from the world of paper publications, perhaps most notably in its interactivity and the fluid nature of the information it contains.
In November 2005, Dan Saffer described Web 2.0 in terms of the experiences associated with it and with an image (Figure 9):
On the conservative side of this experience continuum, we'll still have familiar Websites, like blogs, homepages, marketing and communication sites, the big content providers (in one form or another), search engines, and so on. These are structured experiences. Their form and content are determined mainly by their designers and creators.
In the middle of the continuum, we'll have rich, desktop-like applications that have migrated to the Web, thanks to Ajax, Flex, Flash, Laszlo, and whatever else comes along. These will be traditional desktop applications like word processing, spreadsheets, and email. But the more interesting will be Internet-native, those built to take advantage of the strengths of the Internet: collective actions and data (e.g. Amazon's "People who bought this also bought..."), social communities across wide distances (Yahoo Groups), aggregation of many sources of data, near real-time access to timely data (stock quotes, news), and easy publishing of content from one to many (blogs, Flickr).
The experiences here in the middle of the continuum are semi-structured in that they specify the types of experiences you can have with them, but users supply the content (such as it is).
On the far side of the continuum are the unstructured experiences: a glut of new services, many of which won't have Websites to visit at all. We'll see loose collections of application parts, content, and data that don't exist anywhere really, yet can be located, used, reused, fixed, and remixed.
The content you'll search for and use might reside on an individual computer, a mobile phone, even traffic sensors along a remote highway. But you probably won't need to know where these loose bits live; your tools will know.
These unstructured bits won't be useful without the tools and the knowledge necessary to make sense of them, sort of how an HTML file doesn't make much sense without a browser to view it. Indeed, many of them will be inaccessible or hidden if you don't have the right tools. (Saffer, 2005)
As Saffer says,
There's been a lot of talk about the technology of Web 2.0, but only a little about the impact these technologies will have on user experience. Everyone wants to tell you what Web 2.0 means, but how will it feel? What will it be like for users? (Saffer, 2005)
This idea of versions of the Web is clearly abhorrent to some, as they consider its continuous evolution to be one of its virtues (Borland, 2007), but the significance of the changes in the Web are not denied. These comments are made at a time when there is already talk of Web 3.0. If Web 3.0 represents anything, according to Borland:
Web 1.0 refers to the first generation of the commercial Internet, dominated by content that was only marginally interactive. Web 2.0, characterized by features such as tagging, social networks, and user- created taxonomies of content called "folksonomies," added a new layer of interactivity, represented by sites such as Flickr, Del.icio.us, and Wikipedia.
Analysts, researchers, and pundits have subsequently argued over what, if anything, would deserve to be called "3.0." Definitions have ranged from widespread mobile broadband access to a Web full of on-demand software services. A much-read article in the New York Times last November clarified the debate, however. In it, John Markoff defined Web 3.0 as a set of technologies that offer efficient new ways to help computers organize and draw conclusions from online data, and that definition has since dominated discussions at conferences, on blogs, and among entrepreneurs. (Borland, 2007, page 1)
Kelly (2007) proposes the following, based on ideas from many:
Semantic Web
That internet of things, where everything we make contains a sliver of connection, is still a ways off, although I believe we will create it. The internet of data -- the world wide database -- is quickening right now. As far as I can tell, this is what people mean by the Semantic Web. Because in order to be shared, information is extracted from natural language, reduced to its distinct informational elements, and tagged into a database. In this foundational form it can then be re-assembled into meaningful (semantic) informational molecules in thousands of new ways that is not possible to do when it remains in a flat un-annotated primitive document.I believe this shareable extraction of data is also what people mean by Web 3.0. In this version of the webosphere data surges, flows, and expands across websites as if it were acting within one large database, or within one large machine. My site solicits a steady stream of data from Alice and Bob; it then adds value by structuring the data in a new (semantic) way, and then I issue my own streams of organized data, to be consumed by others as raw data. This ecosystem of data runs on an open transport system, and consensual protocols, even though not all data is shared or public.
An operational Semantic Web, or World Wide Database, or Giant Global Graph, or Web 3.0, will make possible millions of seemingly smarter services. I won't have to re-tell each website who my friends are; once will be enough. If my name shows up in text, it will know it's me. My town will be a town on the web -- a place with definable characters -- and not just another word. That ubiquity enables any references to my town to link to the actual information about the town. The apparent smarter nature of the web will be due to the fact that the web will "know" more. Not in a conscious way, but in a programatic way. Concepts and items represented on the web will point to each other and know about each other -- in a fundamental way they do not right now.
For the purposes of the research, Web 3.0 is defined as a Web in which not only machines but also humans not just contribute to established ‘container sites’ in ways anticipated by the site owners, but they restructure the sites and content and produce their own sites. This scenario includes a world in which users fail to recognize or be concerned with the distinction between their computing environment and the global one. For this to happen, the users are likely to depend increasingly on Web-based software and storage that they personalize, but without necessarily knowing where everything is – just how to get to it and use it.
The research involves recognising and predicting changes. As William Gibson wrote, “the future is here, it is just unevenly distributed” (wikipedia William Gibson, 2006). It is no longer sufficient to work on an outdated model that involves merely electronic publication of traditional materials; the materials have changed and will continue to do so. As the research shows, the evolution of the Web offers both new challenges and new opportunities. Howell (2008) warns:
We need to keep our eyes on web trends and recognise trends that actually help to improve disabled people’s experience of the web. Arguably, personalisation is a trend that actually helps as its focus is on sites’ best possible performance for every user and is a great deal more effective that the ‘one site for all’ approach.
The United Nations Convention (2006) refers to many kinds of digital resources and their location and use without using the word 'Web' despite the recent revolution caused by the development of what is known as the Web, or World Wide Web. Standards Australia, for example, in its 2008 draft metadata standard has included metadata for objects that are not digital, in the following:
This document is an entry point for those wishing to implement the AGLS Metadata Standard for the online description of online or offline resources. (AGLS, 2008, sec.1.1)
They continue:
The aim of the AGLS Metadata Standard is to ensure that users searching the Australian information space on the World Wide Web (including intranets and extranets) have fast and efficient access to descriptions of many different resources. AGLS metadata should enable users to locate the resources they need without having to possess a detailed knowledge of where the resources are located or who is responsible for them. (AGLS, 2008, sec.1.5)
Computer operating systems are now being designed with the user interface driven by metadata in ways that extend the familiar interface of the 'Web' to personal computers and the files within them (for example, Sugar on the XO computer (Derndorfer, 2008), and the Google desktop [Google Desktop].
For this research, the 'Web' is defined as all digitally addressable resources without necessarily distinguishing between the applications or formats in which they are developed, stored, delivered or used by others. This, according to the man credited with the invention of the World Wide Web, Sir Tim Berners-Lee, is 'the Web' and as it develops it achieves more diversified characteristics:
The Semantic Web is an evolving extension of the World Wide Web in which web content can be expressed not only in natural language, but also in a format that can be read and used by software agents, thus permitting them to find, share and integrate information more easily.[1] It derives from W3C director Sir Tim Berners-Lee's vision of the Web as a universal medium for data, information, and knowledge exchange. (wikipedia Semantic Web, 2007)
The essential feature of the Web, then, is that the resource can be addressed; that is, it has a Universal Resource Identifier [URI] that allows it to be found electronically. Such identifier need not be persistent (consistent even for dynamically created content), and the resource need not be maintained in any particular state; it might be constantly changing and it may not have continuity.
Brown and Gerrard (2006) argue that broadband access to Internet makes it easier to make accessible content. This is in line with other expectations for the future; as the technology improves, the opportunities should improve.
It is unlikely that more than 3% of the resources on the Web are accessible (as defined, see Chapter 3). In other words, even if a user has appropriate equipment and has received a resource, the chance that they will be able to perceive the intellectual content of that resource is extremely low if they have special needs. It may be that they have a medically recognised permanent disability such as being blind and the resource is only available as an image of a poem on a tombstone. If so, they may have no idea what it is or what it says. They may have a constructed disability, as a result of driving a car in a foreign country and using their phone to try to get location instructions in a language they understand. The social model of disability (Oliver, 1990b) conflates definitions of disabilities as characteristics of humans and instead adopts the perspective of the human as being disabled by the circumstances, natural or constructed, physical or otherwise (Chapter 3).
(In this thesis, disabilities of a medical nature are described as permanent disabilities. It is recognised that some disabilities naturally increase with age and usually are experienced by all who live long enough.)
The research concerns the accessibility of the Web. Accessibility in this context is a match between a person's perceptual abilities and information or communication technologies and artefacts. Many people have special needs to enable this match, especially people with permanent disabilities. As the UN Convention says:
Persons with disabilities include those who have long-term physical, mental, intellectual or sensory impairments which in interaction with various barriers may hinder their full and effective participation in society on an equal basis with others. (UN, 2006, Article 1)
The use of the term accessibility in this research distinguishes between access as considered in this thesis and access as used to describe possession of facilities for connection to the Web. Nor does it include having the necessary legal rights to use resources. These other kinds of access are, of course, crucial to any user who is dependent on the Web. Such access is often dependent upon socio-economic factors, levels of education, regional and wider factors relating to communications availability and quality, or any number of similar factors. It also may be dependent upon such as intellectual property, state or private censorship, etc. The AccessForAll approach advocated in this thesis is only concerned with access as it relates to users who, for whatever reason, cannot access Web resources, including services, when they are in possession of facilities that should be adequate; in other words, when they cannot access what they already have access to.
This is not an exhaustive definition and will be further elaborated (Chapter 3) but it should be noted that accessibility in this thesis explicitly includes people with conditions that are medically recognised as disabilities and herein described as permanent disabilities.
Metadata is the term used, with reference to the Web, to refer to descriptive labels of content. It is usually applied to descriptions in an agreed format for the creation of machine-readable descriptions of digital resources that can be used for, among other things, the discovery of those resources (wikipedia metadata, 2008; University of Queensland Library, 2008; UK Office for Library and Information Networking, 2008; W3C Technology and Society Domain, 2008).
There is a detailed discussion of metadata in Chapter 6. It explains the multiple uses of metadata and how it comes to be central in the present work.
AccessForAll metadata is, at one level, descriptions of the accessibility characteristics of resources (note that metadata is actually a plural noun, although it is easier to think of it as a singular noun, as is usually done in this thesis). At another level, it is the way in which such descriptions are made. It is metadata at the second, latter level that is the subject of the research. It is described as a metadata framework to distinguish it from the metadata that is content produced according to the framework.
The descriptions enable content providers to create and offer resources that can be adapted to individual needs and preferences. Thus they can minimise the mismatch between people who, especially but not exclusively, have special needs due to permanent disabilities, and resources published within what is here defined as the Web. This is explained further in Chapter 7.
AccessForAll accessibility is based on the use of metadata. By associating AfA metadata with resources and resource components, new services are enabled that support just-in-time, as well as just-in-case, accessibility. Metadata describing individual people's accessibility needs and preferences is matched with descriptions of the accessibility characteristics of resources until the individual user is able to access a resource that satisfies their needs and preferences (Figure 10).
While nothing can prove that the Web will become more accessible or otherwise, this research shows that already there are resources available that could be used to take advantage of AfA metadata, and make the Web more accessible (Chapter 7). AfA metadata was initially for education only. It is being further developed as an ISO/IEC JTC1 multi-part education standard [ISO/IEC 24571] with some parts already completed. It would be very satisfying if AfA metadata was adopted for resources across all domains, especially as a result of being adopted by other standards bodies (already being adopted world-wide by the Dublin Core Metadata Initiative [DCMI] and in Australia as part of the Australia-wide AGLS Metadata Standard [AGLS]).
Metadata specifications, as explained earlier, can be very local. As such, they may only work locally, or they can be shared to form a common vocabulary. Based on earlier similar work in the library world, metadata specifications do not always follow technical development practices. In particular, they are rarely based on a specified model. Practices cannot always be reduced easily to a technical model because there is often inconsistency in those practices.
After ten years of practice, the Dublin Core community have at last managed to state their model explicitly. This has been a difficult process and has involvd fundamental re-thinking of many aspects of the practice that have evolved. Now, given a clear model, all Dublin Core metadata needs to follow that structure. Often, now, it is simply a matter of adding semantics to the model. Similarly, the structure of the metadata for AccessForAll is considered crucial if AccessForAll metadata is to be effective globally. In this research, the particular metadata elements and their potential values are not the focus. These are established by the accessibility community working on AccessForAll, including the author. They are considered to be data used in the research. The detailed specifications containing the relevant definitions are being published (for free) and will be available from various standards bodies' Web sites [IMS GLC; ISO/IEC JTC1; AGLS].
The research provides evidence that there is already metadata available that could be transformed to match an AfA framework, and that other suitable data could be generated automatically from existing data (see Chapter 7). Currently, such metadata is not available for use by those with accessibility needs. As a result, individual users cannot discover, in anticipation of the receipt of resources, if they will be able to access them. The thesis (Chapter 10) explains how individual users, possibly assisted by computer systems, could take advantage of descriptive metadata to meet their accessibility requirements.
In many disciplines, those working within the narrow discourse of a particular discipline, or part of it, tend to use words that can have other or broader meanings in different contexts. The definitions offered above are not exhaustive but are necessary for the reading that follows in setting the context for the research. The research is confined to a small section of information systems work. It faces the problem that some of the terms used, such as accessibility, are easily understood in a general sense by everyone. As such, their particular use in this context can be confusing. What follows further outlines the narrow scope of the research and the context and the methods used in the research.
A significant problem for people with special needs, and their content providers, is that there are often intellectual property issues associated with those resources, especially when they are transformed for access by users. In many jurisdictions, there are special rights for people with recognised disabilities. Their rights may involve complicated intellectual property rules. This is completely beyond the scope of this research that focuses on how such materials can be made discoverable and interoperable, but it is seen as a potential precursor to any work that needs to take place to allow such interaction. (It has been reported informally to the author that the Copyright Agency of Australia is reviewing its practices in this respect following interaction with the AccessForAll work in late 2007).
The research is not about the techniques used to make digital information accessible to people with disabilities. That is the work of the World Wide Web Consortium's Web Accessibility Initiative [W3C WAI]. It is recognized as crucial for those who produce resources because it helps them make content in suitable ways. Without content that is properly formed, the metadata framework proposed cannot help. The research is concerned with how, when a resource is identified as of interest, a user with particular needs at the time and in the context in which they find themselves, can have the intellectual content of that resource presented to them in a way that matches their immediate needs and preferences. If necessary, this includes having components of the original intellectual content replaced or supplemented by the same information in other modes, or having it transformed. The research contributes the potential for this to be done, not the components themselves.
A challenging attribute of digital information is the increasing mobility of people who expect information to be available everywhere, and expect to use all sorts of devices to access it. As they travel from one country to another, users expect to continue to gain information in their language of choice, even though, for instance, it is about places where different languages are spoken. Sometimes users expect to get location-based, or location-dependent, services and sometimes they want location independent services (Nevile & Ford, 2004b, Chapter 3).
The context in which a user is operating is fundamental to the type and range of needs and preferences they will have (Kelly et al, 2008). The research embraces what is known as the Web 2.0 and anticipates Web 3.0. In these Web worlds, an evolutionary progression from the original Web which was created by the technique of referenced resources and distributed publishing, users interact with resources and services that are made available by others, often with no knowledge of their source. (Discussion of the new environment and the way it operates is within scope as it clarifies the context for the work (see Preamble and Chapter 6).)
It should be noted that the W3C WAI work currently considers some Web content out of scope at this stage, in their accessibility work (W3C WCAG 2.0, 2008).
If a resource contains some components that are inaccessible to a user, those components need to be transformed or replaced or supplemented for the user. It is outside the scope of the current research to deal with the problem of discovery of those components or the services that might be used for their transformation. The problem is considered not so much as peculiar to accessibility but as a problem related to modified on-going searching when resources that are discovered prove inadequate. This has been considered recently at the University of Tsukuba, in Japan (Morozumi et al, 2006). It is a topic of research closely related to new work on what are called GLIMIRS (sic) from the Online Computer Library Center [OCLC] in the US (Weibel, 2008a). Understanding of the problem of iterative searching is, however, in scope.
Out of scope also is any requirement to engage with the details of adoption of AfA by industry. Adoption by standards bodies depends upon processes that engage the industry in formal ways, so adoption by such bodies is considered to include adoption by industry. Implementation is, on the other hand, not always ensured by the existence of standards. At the time of writing, before publication of the standards, there are already some significant implementations of the AfA standards. These are discussed in Chapter 11.
A significant aim of the research is the distillation, from a perspective developed from working extensively in the fields of accessibility and metadata, of work undertaken in the accessibility context on what is called the AccessForAll approach to accessibility.
Information science is a well-established field. It brings together computer science and information management, forming a theoretical field for research into information management. The Journal of Information Science describes its field thus (2008):
Information Science is a broad based discipline which has a potential impact in almost every sphere of human activity in the emerging information age.
There have been significant advances in information technology and information processing techniques over recent years and the pace of innovation shows no sign of slowing. However, the application of these technologies is often sub-optimal because theoretical understanding lags behind.
The Journal seeks to achieve a better understanding of the principles that underpin the effective creation, organization, storage, communication and utilization of information and knowledge resources. It seeks to understand how policy and practice in the area can be built on sound theoretical or heuristic foundations to achieve a greater impact on the world economy.
Elsevier (2008) says of their journal, Information Sciences:
The journal is designed to serve researchers, developers, managers, strategic planners, graduate students and others interested in state-of-the art research activities in information, knowledge engineering and intelligent systems. Readers are assumed to have a common interest in information science, but with diverse backgrounds in fields such as engineering, mathematics, statistics, physics, computer science, cell biology, molecular biology, management science, cognitive science, neurobiology, behavioural sciences and biochemistry.
In his essay Information Science, Tefko Saracevic (1999) wrote:
In approaching the discussion of information science, I am taking the problem point of view as elaborated by Popper (1989, p. 67) who argued that:
[S]ubject matter or kind of things do not, I hold, constitute a basis for distinguishing disciplines. . . .We are not students of some subject matter, but students of problems. Any problem may cut right across the border of any subject matter or discipline. (Emphasis in the original)
My emphasis is on problems addressed by information science. Although I provide definitions of information science and other fields in interdisciplinary relations with information science later in the essay, I am doing this solely to advance the understanding of problems addressed by different fields and their relation to information science problems. Debates over the “proper” definition of information science, as of any field, are fruitless, and in expectations naive. Information science, as a science and as a profession, is defined by the problems it has addressed and the methods it has used for their solutions over time. Any advances in information science depend on whether the field is indeed progressing in relation to problems addressed and methods used. Any “fixing,” if in order, will have to be approached by redefining or refocusing either the problems addressed, or the methods for their solutions, or both. (1999, p 1051)
He continues:
A history of any field is a history of a few powerful ideas. I suggest that information science has three such powerful ideas, so far. These ideas deal with processing of information in a radically different way than was done previously or elsewhere. The first and the original idea, emerging in 1950s, is information retrieval, providing for processing of information based on formal logic. The second, emerging shortly thereafter, is relevance, directly orienting and associating the process with human information needs and assessments. The third, derived from elsewhere some two decades later, is interaction, enabling direct exchanges and feedback between systems and people engaged in IR processes. So far, no powerful ideas have emerged about information, as the underlying phenomenon, or “literature” (as defined later), as the object of processing. However, one can argue that the idea of mapping of “literature,” that started with exploitation of citation indexes in 1960s, may also qualify as a powerful idea. (1999, p 1052)
He argues there are two major clusters of work in the field – one around the texts themselves and the other to do with their management, retrieval, storage, etc. He says:
More specifically, information science is a field of professional practice and scientific inquiry addressing the problem of effective communication of knowledge records— “literature”—among humans in the context of social, organizational, and individual need for and use of information. (1999, p 1055)
White and McCain (1998) did extensive data-gathering about the top authors in the field of information science. The placement of names in the map below (Figure 11) represents the connections between the authors. The clusters are, according to Saracevic, on the left around the analytical study of literatures and those on the right more around what has become known as information retrieval. Saracevic notes that the authors are sparsely spread around what he likens to the coastal areas of Australia, signifying little connections with whatever is a core of the field. He notes the sparceness of pure research around literature, as he calls it. He asserts that there is more money for applied research, research with respect to information retrieval, and this partly explains the distribution of authors.
Today, it is safe to guess, one of the ‘coastal towns’ would be occupied by authors writing about metadata. There is a proliferation of journals and conferences in the field, and for all the practitioners, there are many who theorise in universities and research institutes.
Saracevic argues that information science is suffering from being fragmented and that, at the same time, it engages in interdisciplinary work. By this, he means that information science does not operate in a vacuum but rather that it works with many disciplinary fields. He concludes by saying:
I am also convinced that the greatest payoff for information science will come if and when it successfully integrates systems and user research and applications. Society needs such a science and such a profession. (p 1062)
The International Journal of Metadata, Semantics and Ontologies (IJMSO, 2008) describes its objective as follows:
to provide an open forum in which several disciplines converge and provide their perspectives regarding the complex topic of metadata creation, use and assessment. Those disciplines include Digital Libraries, the Semantic Web, Library Science and Knowledge Management, among others.
There is also the Journal of Library Metadata
The journal with the unique focus on metadata applications in libraries. (Beall, 2008)
and a steady flow of new books.
Following arguments such as those presented by Saracevic, based on extensive research of the literature regarding the field of information science, it is to be assumed there is no doubt about its status as a science, and that its strength partly lies in its ability to work collaboratively with other disciplines. A decade later, it is now asserted, there is a branch of information science that has earned the name metadata science, or, at least, that the information science specialty in which the author is engaged, is a metadata cluster.
In 2003, Jane Hunter summarized the key issues and some of the research in the metadata field.
She wrote:
Some of the major disadvantages of metadata are cost, unreliability, subjectivity, lack of authentication, and lack of interoperability with respect to syntax, semantics, vocabularies, languages, and underlying models. (2003, page 1)
Hunter (2003, page 5) distinguishes between metadata research and the development of what she calls ‘upper-level’ ontologies. In the current research, such a distinction is drawn by the split between the work of the AccessForAll team on what Hunter calls the upper-level ontologies, and the work in this research, to do with how to make those ontologies work. The AccessForAll ontologies are contained in the standards, as published, and so are not considered a core part of this thesis.
Similarly, Hunter distinguishes Web services and harvesting. A number of such areas of research are relevant but not the focus of this research. The matching process for AccessForAll, for example, might be considered a Web Service. Here, it is simply considered to exist as a ‘black box’ for which the metadata work is being undertaken. It is not the focus of the work of the research, although its use is a major motivator for the AccessForAll development.
Hunter, citing Clifford Lynch (2001) continues:
The individualization of information, based on users' needs, abilities, prior learning, interests, context, etc., is a major metadata-related research issue ... The ability to push relevant, dynamically generated information to the user, based on user preferences, may be implemented
• either by explicit user input of their preferences;
• or learned by the system by tracking usage patterns and preferences and adapting the system and interfaces accordingly. (2003, page 13)
She wrote:
The idea is that users can get what they want without having to ask. The technologies involved in recommender systems are information filtering, collaborated filtering, user profiling, machine learning, case-based retrieval, data mining, and similarity-based retrieval. User preferences typically include information such as the user's name, age, prior learning, learning style, topics of interest, language, subscriptions, device capabilities, media choice, rights broker, payment information, etc. Manually entering this information will produce better results than system-generated preferences, but it is time consuming and expensive. More advanced systems in the future will use automatic machine-learning techniques to determine users' interests and preferences dynamically rather than depending on user input. (2003, page 14)
For the reasons made clear in this thesis, this is not the way that the accessibility community wants to work, well, not exclusively, It is true that some of the capabilities of the user’s devices should be conveyed automatically to the computer applications involved, but it is essential that there is control left in the user to choose what they want. In this sense, the AccessForAll work is both more than usually dependent on an appropriate enabling technology, such as metadata, and novel, in how it does not simply follow the trend. It is also the case that the AfA work depends much more on a close connection between the technology and very personal needs and practices, the human-interface issues that Saracevic (1999, page 1052) argued are the central and big problems of information science.
Hunter (2003, page 18) concludes :
The resource requirements and intellectual and technical issues associated with metadata development, management, and exploitation are far from trivial, and we are still a long way from MetaUtopia….
The current research is wholly to do with the intellectual and technical issues that have arisen from the exercise of developing AccessForAll metadata. The research methodology for this has been tailored to that task.
Determining issues is never easy. There is inevitably a perspective bias from any written interpretation of data. In the research, the obvious bias is the Dublin Core metadata perspective. This research has been undertaken within a wider context of extensive and broad Dublin Core metadata activities over more than a decade (Appendix ???). Here, the bias is embraced not avoided: Dublin Core metadata is arguable the most prolific and it is the recommended metadata in the Australian context [AGLS]. Since the early days of metadata in the context of the Web, and the use of the Web for publishing by them, the Australian governments and their cultural institutions have supported the use and development of Dublin Core metadata. The research has challenged aspects of Dublin Core metadata without fear or favour, and in some cases contributed to its clarification through the research (Pulis & Nevile, 2006). The bias is, then, not blind in favour of the Dublin Core work but certainly towards Semantic Web technologies as fundamental to the potential of metadata, and thus Dublin Core metadata as the most appropriate expression of metadata for the Semantic Web. There is an explicit effort to make DC metadata Semantic Web interoperable, and this has been of interest since the early days when the PICS protocol was first being used and the Semantic Web was developed as an extension of that. (Lassila, 1997)
The research is multifaceted and so involves a number of strategies. The aim is to use the different strategis appropriately to inform the writing of an integrated theory of the metadata that can support the AccessForAll approach to accessibility. The demands of that approach have been used as drivers for the problem-solving that Saracevic (1999, p. 1051), by implication, suggests will define the science of metadata.
The research does not include any surveys or quantitative data, or analysis of original data, although it references extensively work that does this, and provides some critical interpretation of those works. It is theoretical work that attempts to make sense of what can and should be done to achieve a set of goals. The goals are examined in detail because they need to be deeply understood for the work to be done. As argued above, this is not repetitive work. It is not about building an ontology for accessibility but about how such an ontology should be built, and why. The building of the actual ontology is critical grist for that mill. The details of the work are referred to time and again to show exactly how they should be understood and responded to.
Pioneering research (Seely Brown, 1998) is about research that builds something new, a new technology, that cannot be tested other than by time. This is not such research although it shares some characteristics.
John Seely Brown (1998) differentiated between what he thought of as two main kinds of research, sustaining and pioneering. Sustaining research, he thought, is aimed at analysis and evaluation of existing conditions. The problem for researchers in fast-changing fields is that often, by the time sustaining research is reported, the circumstances have changed. As the original circumstances cannot be reproduced, the research results need to be interpreted into a different context to be useful and in some fields, this cannot happen. In the case of pioneering research, the work is successfully implemented or, perhaps more often, forgotten. This is the sort of work in which many technology focused researchers are engaged: they follow what are traditional research practices to a point, but their work is evaluated differently and they need to engage with and accept different types of evaluation.
It is not possible to test today if the AccessForAll metadata will make a difference. At best, it can be estimated what potential it has for doing this, and this should be done in an informed way. But the research is not about a product that has to be successful, it is about how to make such a product, what are the relevant factors. Similarly, it is not about what has been done in the past when a metadata project has been successful, although that is relevant. There are lessons from the past that are taken to inform the work, but this is a new type of metadata for a new purpose in a new technological context. It is not, therefore, the purpose of the research to determine what has made a particular metadata development successful although the factors that have done this in a number of cases should be identified and articulated and their contribution evaluated. This has been done.
What has been happening is that, in the context of working on accessibility and ontology development, the principles and practices of the field have been exposed and they have been identified, analysed, and evaluated. This has happened in the light of extensive reading of the literature and relevant practical experiences. The ‘field’, in this case, has been at the junction of the fields of accessibility and metadata research.
In the field of accessibility, almost all effort has focused on a single set of guidelines [WCAG-1] with what this thesis argues are less than satisfactory results. It is important to evaluate AccessForAll accessibility to ensure this does not happen again. For whatever interest there is in the idea of AccessForAll metadata, there is still a need for research to discover how to create a suitable awareness of the context for the work and the value of the work. This means developing a strong understanding of the theoretical and practical issues related to accessibility, including practical considerations to do with professional development of resource developers and system developers, and the administrative processes and people that usually determine what these developers will be funded to do. It also involves the reading and writing of critical reviews of other work. In particular, while there is little doubt of the potential benefit to users with disabilities, it is not at all clear how to work with the prototyped AfA ideas to make them mainstream in the wider world, both in the world outside the educational domain and in the world of mixed metadata schemas (correct use of this word would be schemata but common usage accepts schemas).
Many use the expression 'research and development' to differentiate between research and development. Development work is so characterised without regard for the processes involved in achieving it. One is reminded of Mitchel Resnick's story of Alexandra whose project to build a marble-machine was rejected as not scientific until the process was carefully examined and she was awarded a first prize for the best science project (Resnick, 2006). In some fields, development is not about something that can be controlled over time. It is about repeatedly designing, creating, testing, evaluating and reviewing something in an iterative process, often towards an unknown result but according to a set of goals. These are also important processes for research. Such processes benefit from rigorous scrutiny that can be attracted in a variety of ways, including by being undertaken in a context where there are strong stakeholders with highly motivated interests to protect.
Designers of educational computer environments, such as Andrea diSessa, research in a field they call ‘design science’ (diSessa, 1991). The ‘experiments’ carried out by such researchers were first designated ‘design experiments’ by Allan Collins (1992). They are used to support the design of environments, or what Fischer calls meta-design. (Fischer & Giaccardi, 2006)
Ann Brown (1992) describes the problem of undertaking research in a dynamic classroom. As an accomplished experimental researcher, she argues that it is not possible or appropriate to undertake experimental research in a changing classroom.
Collins later wrote (1998):
Historically, some of the best minds in the world have addressed themselves to education; for example, Plato, Rousseau, Dewey, Bruner, and Illich. But they addressed education essentially as theorists, even when they tried to design schools or curricula to implement their ideas. Today, some of the best minds in the world are addressing themselves to education as experimentalists. Their goal is to compare different designs to see what affects what. Technology provides us with powerful tools to try out different designs so that, instead of theories of education, we can begin to develop a science of education. However, it cannot be an analytic science, such as physics or psychology, but rather a design science, such as aeronautics or artificial intelligence. For example, in aeronautics the goal is to elucidate how different designs contribute to lift, drag, and maneuverability. Similarly, a design science of education must determine how different designs of learning environments contribute to learning, cooperation, and motivation.
Collins says, “in aeronautics the goal is to elucidate how different designs contribute to lift, drag, and maneuverability”. The equivalent in the current context might be: "to elucidate how different designs contribute to proliferation, interoperability, effectiveness and user experiences". The particulars in focus in this research are the effectiveness and interoperability as there is not yet enough of the AfA metadata to determine how it will affect user experience and the proliferation cannot be pre-determined.
In "Design-based research: An emerging paradigm for educational inquiry",
The authors argue that design-based research, which blends empirical educational research with the theory-driven design of learning environments, is an important methodology for understanding how, when, and why educational innovations work in practice. Design-based researchers’ innovations embody specific theoretical claims about teaching and learning, and help us understand the relationships among educational theory, designed artifact, and practice. Design is central in efforts to foster learning, create usable knowledge, and advance theories of learning and teaching in complex settings. Design-based research also may contribute to the growth of human capacity for subsequent educational reform. (DBRC and D.-B. R. Collective, 2003)
The complexity of the accessibility work is like that of education: everything is constantly changing, including the technology, the skills and practices of developers; the jurisdictional contexts in which accessibility is involved and the laws governing it within those contexts, and the political environment in which people are making decisions about how to implement, or otherwise, accessibility. There are also a number of players, all of whom have different agendas, priorities and constraints, despite their declaration of a shared interest in increasing the accessibility of the Web for all.
The main idea behind design-based research is that in the process of design, one makes explicit the issues that are relevant, and their context, so these can be dealt with by the researcher. The actual design is of something that can be thought of as the motivation for the activity, but is not the product of the research. The research is design-based, not design. The design informs the research.
The Australian Research Council funded the Clever Recordkeeping Metadata (CRKM) Linkage Project in 2003-2005 (ARC, 2007). It was a major metadata research project for Australia and so the research methods used are of interest. The project reported:
The first iteration of the CRKM Project investigated a simple solution for demonstrating the automated capture and re-use of recordkeeping metadata. The expectation was that this initial investigation would expose the complex network of issues to be addressed in order to achieve metadata interoperability and automate the movement of recordkeeping metadata between systems, along with enabling researchers to develop skills and understandings of the existing technologies that support metadata translation and transformation. (CRKM, 2007)
The project demonstrated the use of an established computer systems development methodology in the metadata context. In this case, what the researchers did might be described by some as follow a computer systems development methodology, but by others as design-based research. In this case, the Australian Research Council (ARC) did not fund the development for the value of the development, per se, but because the process of development would inform the research. The current research is in a similar position. In 2002, the author was the principal Investigator in a major ARC project that, at that time, broke all the ARC rules by funding development (ARC, 2001). The ARC accepted the argument that only by doing the development, could the lessons be learned that were essential to the research.
In the introduction to a paper describing the research methodology for the CMKM project, Evans & Rouche (2006) claim:
Systems development research methods allow exploration of the interface between theory and practice, including their interplay with technology. Not only do such methods serve to advance archival practice, but they also serve to validate the theoretical concepts under investigation, challenge their assumptions, expose their limitations, and produce refinements in the light of new insights arising from the study of their implementation. (p. 315)
Engagement with the development of AccessForAll metadata enabled accessibility research that "needs to encompass methods that investigate how emerging theories are operationalized through systems development". In the case of the CRKM project, the researchers were interested in discovering how schemas played a role in the archival context so they would know how to build a metadata registry that uses such schemas (p. 316). In the current accessibility research, the focus is on how to develop metadata schemas for use in content discovery, matching and delivery systems, to improve the accessibility potential of the Web.
The CRKM registry was to provide content for use in a harmonisation of schemas to inform a standardisation process. In the words of the researchers:
The purpose of such a registry of metadata schemas is to act as a data collection and analysis tool to support comparative studies of the descriptive schemas.
With no existing blueprint for such a registry, the first task of the research team was to conceptualise the system and establish its requirements. In so doing several key questions are raised including: – What are the salient features of metadata schemas that need to be documented for the purposes outlined above? How are these realised as elements? ... In order to address these questions, the research team looked at utilizing systems development as an exploratory research approach. (p. 317)
Attention is drawn to the expression ‘exploratory research approach’ here. When research is being undertaken to elucidate ill-defined issues, exploration of them is what, in fact, is required.
In some research projects, design-research is undertaken in formal, tight, iterative phases. Something is designed, built, tested, redesigned, in a number of phases. At other times, however, design research is undertaken in less linear cycles, with these processes often happening simultaneously. So it is in this case. The strategies adopted in the research are those that are expected to best elucidate what is required of the metadata. Although these are exercised simultaneously in most cases, they have been reordered for reporting purposes into what is now more or less a sequential set of activities.
The research provides the first significant description of issues relevant to AccessForAll metadata and how it can be used. It justifies the development of such an approach to accessibility, and shows how the actual metadata schemas can be developed. This involves a wide range of research activities, as shown below.
To investigate how effective accessibility efforts were in a typical organisation, the author was involved in the auditing of a major university Web site (Nevile, 2004). The process was significantly simplified by the combined use of several automation tools. The audit produced descriptions (metadata) of the accessibility characteristics of the 48,084 pages reviewed.
To facilitate the use of the WCAG specifications by content developers, the Accessible Content Development Web site (Appendix 8) was built. The aim was to provide a fast look-up site organised by topic and focus, rather than the lengthy, integrated approaches required at the time by anyone using the W3C Web Content Accessibility Guidelines [WCAG-1]. As a result of doing this work, the author gained a more structured view of the difficulties being tackled by developers in practice. This complemented previous work in which the author had, on many occasions, been consulted with respect to building accessible sites or to ascertain the accessibility or otherwise of sites, and many times commissioned to repair the sites.
To develop an automatic service for conversion of MathML encoded mathematics into Braille, a major Braille project was undertaken at La Trobe University with support from Melbourne University and the Telematics Trust Fund [Telematics Trust]. The first task was to understand the problems, then to see what partial solutions were available, and then to develop a prototype service to convert mathematics texts to Braille. In this case, there was no need to survey anyone to determine the size of the problem or the satisfaction available from existing solutions - the picture was patently bleak for the few Braille users interested in mathematics and, in particular, the text was required by a Melbourne University student for his study program. Ultimately, the research was grounded in computer science, where it is common to have a prototype as the outcome with an accompanying document that explains the theoretical aspects and implications of the prototype. In this case, the prototype encoding work was undertaken by a student who was supervised by the author, who herself managed or personally did much of the other work in the project (Munro, 2007; Nevile et al, 2005).
To gain an insight into formal empirical research documenting specific problems with the W3C WAI Web Content Accessibility guidelines, the author studied the UK’s Disabilities Rights Commission’s review of 1000 Web sites. This was the first major review of Web sites that evaluated the WCAG’s effectiveness. Many of the findings have more recently been substantiated in other work (see Chapter 4) and they have been anecdotally reported by the author and others for many years.
The author wanted to know that AccessForAll metadata could be applied automatically to resources of interest for their accessibility, using their existing metadata descriptions. So information from major suppliers of accessible resources was gathered to verify the existence of such resources and their metadata (Chapter 7).
To learn how AccessForAll ideas might operate in a distributed environment, the author studied the Functional Requirements for Bibliographic Records (FRBR) and associated work and tried to determine how resources should be described so that other resources with the same content but represented in different modes or with other variations might be discoverable. This work was undertaken at the Knowledge Communities Research Center [KCRC] in Japan with colleagues who, at the time, were trying to learn from FRBR and the OpenURI work. The author is more inclined to think that a new approach to resource description to be known as GLIMIRs may, in fact, prove more useful in this context (see above).
To make sure AccessForAll metadata would be interoperable with other metadata systems, and DC metadata in particular, the author studied the emerging DC abstract data model. To this end, the author worked with data models expressed in formal notation (Unified Modeling Language [UML]). Through this work, the author discovered the ambiguity of the DC Abstract Model as first expressed and became involved in work to clarify that model (Pulis & Nevile, 2006). Eventually, the DC model was expressed in UML and the model proposed for the DC implementation AccessForAll metadata was matched to that model. There is a strong feeling emerging that unless data models are matched, the metadata cannot interoperate without a significant loss of data (Weibel, 2008b).
There are many major players in the field of accessibility. These stakeholders had to be won over. There is really no other way that technologies such as metadata schemas proliferate on the Web. Without the engagement of major players, the technologies are not useful, as explained above. 'Winning over' bodies that use technologies usually means providing a strong technical solution as well as compelling reasons (in implementers' eyes) for adoption of those technologies. In the case of accessibility metadata, the technical difficulties are substantial. As explained in the section on metadata, there are many kinds of metadata and yet they share a goal of interoperability - essential if the adoption is to scale and essential if it is to be across-institutions, sectors, or otherwise working beyond the confines of a single environment. The problems related to interoperability are considered later (Chapter 11) but they are not the only ones: metadata is frequently required to work well both locally and globally, meaning that it has to be useful in the local context and work across contexts. This tension between local and global is at the heart of the technical challenges for adoption when diverse stakeholders are involved and so competes with the political and affective challenges.
At the time the AccessForAll work was being undertaken, a major review of accessibility was being undertaken by the ISO/IEC JTC1. A Special Working Group [SWG-A] was formed to do three things: to determine the needs of people with disabilities with respect to digital resources, to audit existing laws, regulations and standards that affect these, and to identify the gaps. Concerned that this was merely a commercially-motivated use of a standards community with an agenda to minimise the need for accessibility standards compliance, the author asked to know the affiliations of the people represented in the Working Group. Most were employed by one single, major international technology company although they were presented as national body representatives. There were very few representatives of disability or other interest groups. In fact, when the author asked if the people present could identify their affiliations, it took an hour of debate before this was allowed. Not only was the author uneasy about the disproportionate commercial representation, but it emerged that the agenda was constantly under pressure to do more than the stated research work, and to try to influence the development of new regulations that were seen to threaten the major technology companies. Although heavy resistance to the 'commercial' interests was provided by a minority, and in the end the work was limited in scope to the original proposals, it showed just how much effort is available from commercial interests when they want to protect their established practices. Given that many of the companies represented in the SWG-A are also participants in consortia such as W3C, IMS GLC, it is indicative of what was potentially constraining of the AfA work of the AfA team. More recently, this trend has again been demonstrated by the efforts of Microsoft to have their proprietary document standard OOXML approved as an international standard. In that case, there have been legal cases about the problems of representation and decision-making (McDougall, 2008).
Finally, simultaneously with the research, the ISO/IEC JTC1 SC 36 WG 4 has been developing a standard for metadata for learning resources [ISO/IEC N:19788]. The author has been active in this work as it will result in a standard to be followed that should avoid the problems revealed by the research in the case of the AccessForAll metadata.
In design experiments, or research using design experiments (design research), it is a feature of the process that the goals and aspirations of those involved are considered and catered for. In fact, as the work evolves, the goals of the various parties are constantly revisited as the circumstances change and the research informs the design of the experiments.
The current research is not about a researcher testing a hypothesis on a randomly selected group of subjects; the stakeholders and the designers interact regularly and advantage is taken of this to guide the design. The practical aspects are constantly revised according to newly emerging theoretical principles and the new practical aspects lead to revised theories. The goals do not change but the ways of achieving them are not held immutable.
In this work, considerable interaction occurs between the author as researcher and the author and colleagues and other stakeholders in the design process. This is especially exemplified in the various voting procedures that move the work through the relevant standards bodies. These formal processes take place at regular intervals and demand scrutiny of the work by a range of people followed by votes of support for continued work. Challenges to the work, when they occur, generally promote the work in ways that lead to revisiting of decisions and revising of the theoretical position being relied upon at the time. Such challenges also provide insight for the researcher into the problems and solutions being proposed.
In particular, the author sits between two major metadata camps. Those involved in the IMS GLC have experience mainly with relational databases and LOM metadata, which is very 'hierarchical'. On the other hand, the DC community is biased towards 'flat' metadata. The DC view inevitably influenced the author, given her role as Chair of the DC Accessibility Working Group (later the DC Accessibility Community) and membership of the Advisory Board of DCMI. The IMS GLC's interests are for an outcome that will suit them but, as the author saw it, could risk even further fragmentation of the total set of resources available to education, and so not serve the author's real goal which is to increase the accessibility of the Web (of resources).
During the research period, DCMI itself has been wrestling with the problem of interoperability of the LOM and DC educational community's metadata, a difficulty that has been present since the first educational application profile was proposed by the author nearly a decade ago (for Victorian Education Channel). The interoperability is necessary given that, for example, government resources might be used in educational settings and if their metadata can not be cross-walked (see Chapter 6) from one scheme to the other, the descriptions of the government resources will not be useful to educationalists, which seems ridiculous. One way to ease the problem might have been to develop a standard that suits both metadata systems and that might have been possible (see Chapter 11), but there was insufficient technical expertise available to achieve that goal, so the best that could be done in the circumstances became the modified goal. This has been achieved and it is possible to cross-walk between the various metadata standards so that it does not matter so much which is used, because the data of the metadata descriptions can be shared without loss.
The design work reported has been progressively adopted and has now become part of the Australian standard for all public resources on the Web (as the AGLS Metadata Standard) and by virtue of being an ISO standard [ISO/IEC JTC1 N:24751], an educational standard for Australia. This can be taken as indicative of it having proven satisfactory to a considerable number of people. Only actual implementation and use will prove it to have been truly successful because it will need to be proliferated to the extent that it becomes useful. Implementations are discussed further in Chapter 11.
In this chapter, attention has been paid to the immediate need for working definitions of the terms accessibility, metadata, Web 2.0, and others, and the scope and type of the research. These will all be further elaborated in the following chapters.
The research establishes that, given an understanding of the field of accessibility, the context for it, and frustration with the lack of success and the results of recent research, it is evident that for all the good intentions, there has been poor implementation of accessibility techniques. Universal design is not a sufficient strategy even if it is applied, and a narrow focus on specifications for authoring of Web content alone will not produce the desired results. This means there is a need for a new approach. By using a range of existing and emerging standards and introducing metadata to describe user needs and preferences, it is possible for them to be matched by resource characteristics, also described in metadata. By adding this possibility, without compromising interoperability of metadata or stakeholder interests, and by attracting implementation, individual access needs and preferences should be able to be satisfied. This AccessForAll approach places emphasis on the accessibility of the Web for individuals, and involves many standards working together. It does not depend upon universally accessible resources but includes them. It will be elaborated in the next chapter.
