[Accessforall] Draft of new standard

Andy Heath andyheath at axelrod.plus.com
Tue Jul 3 09:38:51 UTC 2012

I've added a couple of point to Liddy's below. -andy
> Thanks Christobe,
> my comments in situ.
>> (1) The scope section states: "ISO/IEC 24751 is intended to facilitate
>> the matching of resources to user's individual requirements by
>> providing common terms for describing those requirements so they can
>> be matched to resources that satisfy the requirements." But what about
>> groups of users? The approach taken by the VUMS cluster does not
>> assume that a user profile can only belong to an individual user: "A
>> user profile is an instantiation of a user model representing either a
>> specific real user or a representative of a group of real users" (from
>> the VUMS Cluster white paper:
>> <http://www.veritas-project.eu/vums/wp-content/uploads/2010/07/White-Paper.pdf>).
> It is, I believe, made clear that a profile might be only one of many
> that a single user, in fact, uses throughout their time. Similarly, it
> is expected that a number of users with similar needs will, at times,
> all use a single profile. I will try to make sure this is made clear.
> One of the problems, of course, is that a set of users may have
> conflicting needs but that is not your question, I realise.
>> (2) The registry is supposed to contain terms that enable the
>> description of needs and preferences, as opposed to medical or
>> functional characteristics of users. However, will Part 1 and/or the
>> conformance clause state that terms for medical or functional
>> characteristics make the implementation non-conforming? The current
>> draft does not prohibit terms for medical or functional characteristics.
> Our work has very deliberately not supported any description of people -
> their needs are not considered definitive of them, etc., and we don't
> support descriptions such as: resources for a blind person. This
> approach is based on the belief that as people are so different in their
> skills and needs, it actually does not make sense to do this anyway.

I completely agree with Liddy's points and would not personally support 
any use of a profile describing medical needs or based on some 
description of people.

With respect to the matter of whether a profile could support a group of 
people there are useful use-cases where it could. For example:

1. The BBC MyWay has initial profiles where each has been discovered to 
be useful by a number of people.  The idea is a user can get going 
quickly by picking one (say one for vision impairment) then modify it to 
more closely meet their individual needs as they go along.  These are 
based on CSS not on the level of preferences in 24751 but the idea is 
similar and could work with 24751 preferences.

2. In the EU4ALL project we looked at procurement of learning resources 
(i.e. specific media - say captioned video) to meet the needs of 
students on a course as expressed by their PNP's.  I don't remember how 
it was implemented but I believe the matching was done within the system 
so that a list of required resources came out but its perfectly possible 
that a system might construct a PNP that doesn't relate to one user but 
to a group of users and represents the set union of their requirements 
and pass this to an organisation that supplies services, modifies 
resources etc.  It would be pretty useful to be able to check a course 
against such a profile to determine whether it can meet its 
requirements.  The interpretation of any priorities in such a PNP might 
be somewhat different from an individual PNP and its possible that there 
might be different structural requirements in the way a PNP is 
represented since two particular requirements might relate to different 
users. In view of this, it might be preferable (and my view is it *is* 
preferable) to consider such an object not as a PNP at all but something 
else beyond our scope, and to take the view that a PNP relates only to 
an individual, but could be a generalised one (as the bbc one described 
above is - at any one time it relates to one user) in some circumstances.

> Liddy
>> From: Liddy Nevile <liddy at sunriseresearch.org>
>> To: Gottfried Zimmermann <zimmermann at accesstechnologiesgroup.com>
>> Cc: "Accessforall at fluidproject. org" <Accessforall at fluidproject.org>
>> Sent: Thursday, 14 June 2012, 5:12
>> Subject: Re: [Accessforall] Draft of new standard
>> I am not sure if anyone else is doing something but I have done more
>> work on the draft of the revised 24751 that is on the wiki at
>> http://wiki.fluidproject.org/display/ISO24751/New+Version+of+24751+Part+1
>> Liddy
>> _______________________________________________
>> Accessforall mailing list
>> Accessforall at fluidproject.org
>> http://lists.idrc.ocad.ca/cgi-bin/mailman/listinfo/accessforall
> _______________________________________________
> Accessforall mailing list
> Accessforall at fluidproject.org
> http://lists.idrc.ocad.ca/cgi-bin/mailman/listinfo/accessforall


Andy Heath

More information about the Accessforall mailing list