Re: essential use case of personlization missed out

Hi Alistair and David

There are open source symbol sets. There are also ways to reference open source node sets . 


There are different ways to link the the symbol to a concept node such as using linked data concept codes or the simpler cogs semantics. there was a browser that changed the symbol set based on the concept code used, and the users preferred symbol set, it is opensource on github. But there was not the content to support it. 


We tried to outline the issue and techniques examples in an  issue paper (https://rawgit.com/w3c/coga/master/issue-papers/symbols-non-verbal.html)


What we need to do is have a requirement that content made for this audience uses some form of linked data on the symbols so that different symbol sets can be loaded that the user is familiar with. It is really enabling personlization of symbols, but not just symbols in controls  -but on symbols used as language. the scope really needs to be written in terms of audience not content - and that is the challenge



All the best

Lisa Seeman

LinkedIn, Twitter





---- On Wed, 26 Jul 2017 14:27:59 +0300 Alastair Campbell<acampbell@nomensa.com> wrote ---- 

      Thanks Joe,
  
 So if I understand, the requirement is for every site to:
  Write all content at lower than a secondary education level, possibly targeted to a particular symbol set.
Install a (possibly proprietary) symbol set plugin.
  
 The second part is hugely problematic, especially as you mentioned different people might want/require different symbols sets, not to mention the immediate and obvious cost overhead.
  
 Putting that to one side and just looking at the reading level aspect, that also has huge issues:
  It would mean essentially promoting a current AAA SC to A, despite the fact it was at AAA for good reasons (which are probably the following).
It cannot apply to all content. For example, we have a client that publishes Physics papers.
It introduces a huge burden, the parallel with BSL is a good one and that is not required by WCAG either.
  
 Unless there is a way to narrowly scope a version that helps without triggering the above issues (e.g. applying to labels & instructions only, but I thought that was there already?), I don’t think it is ready to progress.
  
 I would have thought the steps for progress would need to be something like:
  Establish a non-proprietary symbol set that can be used without license fee.
Establish a browser plugin (or equivalent) that can apply that symbol set (or multiple ones) to any content.
Detail the content-requirements that enable the symbols to be automatically applied.
Form an SC based on the content requirements.
  
 If those first three items have been done, great, but that isn’t the case from what you and Lisa have said.
  
 -Alastair
  
  
  From: Joe Chidzik 
 
   
 
 Addendum to my previous comment: the browser plugin I’m aware of is actually a site level plugin, not at the browser level. So the site would provide the symbolised content via some necessary JS\CSS, and the user would enable it if they want (it requires no plugin). This would mean that the site would have some responsibility to provide content that translated well into symbols. It works by the user enabling the feature, and then hovering the mouse over words to see the equivalent symbol\s)
  
    From: Joe Chidzik 
 
 
  
 I have some experience of the use of symbols to support people with cognitive\literacy difficulties. 
  
 A key aspect of providing symbolised versions of content is that the content is re-written specifically to be translated into symbols. As Lisa points out, there are multiple symbol libraries which people use. 
  
 Providing a direct translation of a portion of text into its equivalent symbol language (Rebus, Makaton, BSL etc) would not provide a satisfactory outcome; in much the same way as machine translating from one language to another verbatim will have undesirable results. Having a symbolised document littered with common symbols (a, the, and) does not necessarily make it more readable. Like BSL, there must be some thought to the grammar and language structure of symbol documents.  
  
 Not sure what the solution would look like, but suspect it would be partially addressed by the current AAA SC 3.1.5 Reading Level: https://www.w3.org/TR/UNDERSTANDING-WCAG20/meaning-supplements.html - Having content written that is easy to understand, and avoids jargon\terminology, would have better results at being translated into symbols. 
  
  
 I do know of at least one proprietary browser plugin that provides symbol support on a per-word basis for users; this functions broadly in the same way as BrowseAloud, but in this case, it is the end user who pays for the product, not the site owner. This at least allows users to get different levels of support, but doesn’t take into account content on sites being written to take advantage of these symbols. 
  
  
 Joe
  
    From: lisa.seeman 
 
 
  
  Content that is made specifically for symbol users uses symbols by most of the text or in place of it- not just controls.
 
 We can add this to the AAA level but we really urgently need it a A or AA so that people can talk to each other and symbol users can share content.
  
  All the best
 
 Lisa Seeman
 
 LinkedIn, Twitter
 
   
 ---- On Wed, 26 Jul 2017 10:15:04 +0300 acampbell@nomensa.com wrote ----
 
 
 
   Hi Lisa,
                                                                                                                           
 I’m a bit confused, I take it this is separate from the purpose of controls AA one we’re discussing, and the AAA equivalent of that.
  
 In which case, what is the content requirement? What is it sites would need to do to support symbol use?
  
 Cheers,
  
 -Alastair
  
   From: lisa.seeman
 
 
  
  Hi Folks
   
 
  We need to see a single A requirement interoperable symbol set codes for non-verbal people. Products for people who are non vocal often use symbols to help users communicate. These symbols are in fact peoples language and people spend years learning a set of symbols tat can not be used with content that is made for these users but by a different company or indeed talk to other people with the same disability but who have learnt the symbols from a different company. Unfortunately many of these symbols are both subject to copy write and are not interoperable. That means end-users can only use one device, and can not use apps or AT from a different company. An open set of references for symbol codes for these symbol sets however, could be interoperable. That means the end user could use an open source symbol set or buy the symbols and use them across different devices or applications. Symbols could still be proprietary but they would also be interoperable. 
   
 
  We need to include this as soon as possible. In the mean time people can lose their language for copy write reasons. However with the current personlization proposal it is only in at AAA. It is such a basic infringement on human rights that it needs to be single A (or higher - this is peoples ONLY way to communicate)
 
 We can make a supporting technique to 1.1.3 , but the "or text" clause plus the requirement of 1.1.1 makes it a redundant technique. 
 
  
 Can we try and find some way to get this in as single A? Technology support includes concept codes or linking to a wordnet concept
 
 
 Lisa Seeman
 
 LinkedIn, Twitter