Online demonstrator for accessible content processing
The EUAIN Demonstrator was launched in February 2006 and the initial reactions were very positive. The Demonstrator was set up in order to illustrate the potential of accessible publishing and can be used for producing different output formats on-demand, from the same well-structured input file. It was emphasised however that this demonstrator offered limited functionalities so that it cannot be considered as a tool to be used for a production purpose. This was an important qualification as a number of publishers have expressed interest in using such a system.
The XML Digital Talking Book (DTBook) format (NISO Z39.86) (also known as the DAISY format) was chosen as a pivotal format on which converters are applied on the fly to produce the output documents in different possible formats (HTML, PDF). Moreover, in order to make the process accessible to a wider audience and to demonstrate that well-structured documents can be produced easily with standard word processors, i.e. by non-specialists, input files can also be provided in an Open Document Format (.odt files). The Open Office word processor can be used for that. The use of Open Document Format is also important as this format is being increasingly recognised as pivotal for public sector information provision.
By way of illustration, the OpenDocument Format (ODF) is to be the standard format for exchanging documents within the public sector in Belgium. This work has been recently strengthened by the launch of the Open Doc Society based in the Netherlands and many countries are now working in this direction.
The Demonstrator has a full user guide and instructions and provides a a set of sample documents to undergo conversion to accessible formats (see diagram below). In addition, the Demonstrator also provides a document structure checker operating on XML DTBook documents. For a given document, this structure checker generates a report in HTML giving an overview of the structure of the document. If the input document is not an XML DTBook document, this checker uses the converters of the multichannel publishing chain to create a DTBook and analyses it.
The generated report shows:
- the metadata included in the document (title, language, isbn, ...)
- the full table of contents of the document structured as a tree
- how many page numbers are tagged in the document and the first and last value found
- for each image referenced in the document, if it is associated with a caption , a short alternative and one or several production notes
- for each table of the document, how many columns and rows it contains, if it contains a caption and headers
- how many footnotes, endnotes, lists, side bars, external and internal links are present