1. This is a great step forward in being able to encode accessibility
information with metadata and Pete should be congratulated.
2. Ed Barker's comments are very valid. If accessibility metadata is to
be widely used then it will need to fit with the IEEE LOM as closely as
possible. Any requirement for extensions is a reason for a tool
developer not to include it, particularly if the community of users is
not yet defined. If this can be incuded in LOM, as Ed suggests, under
"classification" it will be taken up much more readily by those
developing tools because those that suppot LOM will automatically
support this.
3. I have strong reservations about the use of numbers to reflect degree
of accesibility. I think descriptiive terms should always be used for
several reasons:
- when creating metadata, each person will have a different view of
where the boundaries lie between different numerical levels and
inconsistencies will result. Descripotive terms shouyld be chosen such
that they help differentiate levels.
- when reading metadata, numbers alone convey no information and it
would be necessary to lookup some additional table of information to
understand what is meant. If the explanation is included with the number
then the number is irrelevant. Numbers may be used as taxon ids, as Ed
has pointed out, but most users do not wish to see these.
4. The sections which appear as notes (ie "Taking into consideration")
could provide more detailed classification, so for example the section
AccessibleLanguage, which Ed defined as 5 in the classification, could be:
5 AccessibleLanguage
5.1 Use of structural mark-up
5.1.4 - Totally Accessible
5.1.3 - Mostly Accessible
5.1.2 - Partly Accessible
5.1.1 - Barely Accessible
5.1.0 - Inaccessible
5.2 Semantic density
5.2.4 - Totally Accessible
5.2.3 - Mostly Accessible
5.2.2 - Partly Accessible
5.2.1 - Barely Accessible
5.2.0 - Inaccessible
5.3 Quality of writing
5.3.4 - Totally Accessible
5.3.3 - Mostly Accessible
5.3.2 - Partly Accessible
5.3.1 - Barely Accessible
5.3.0 - Inaccessible
Some people will argue that this is too great a degree of complexity. I
think it gives the option for highly detailed definition but does not
stop people from using more "broad-brush" classification. For example,
there could also be a classification
5 AccessibleLanguage
5.0.4 - Totally Accessible
5.0.3 - Mostly Accessible
5.0.2 - Partly Accessible
5.0.1 - Barely Accessible
5.0.0 - Inaccessible
which could generally describe the accessible langauge without going
into the specifics of each aspect. The level of detail to include would
be at the discretion of the metadata creator.
5. I don't think the terms "Totally Accessible, Mostly Accessible,
Partly Accessible, Barely Accessible, Inaccessible" are sufficiently
distinguishable and it may be better to assume a 3-point scale rather
than a 5-point scale.
Charles
----------------------------------------------------------------------
Intrallect Ltd [log in to unmask]
The Glass Cube http://www.intrallect.com/
Houstoun Road, Livingston, t: +44 870 234 3933
EH54 5BZ, Scotland f: +44 1506 498 073
|