See also: IRC log
<MichaelC> trackbot, start meeting
<trackbot> Date: 12 November 2013
<MichaelC> scribe: MichaelC
dd: MMI thinking of writing Note on multi-modality for accessibility
speech, graphical, how multi-modality affects a11y
for cognitive, EmotionML is interesting
nearly Rec
use cases for people who have trouble recognizing emotions
tools to help learn to recognize
use cases:
1) representing emotions that have been recognized e.g., via face recognition, gait, voice, intense tapping on keyboard, etc.
2) synthesizing emotions e.g., in speech synthesis
put together, could express an emotion in input and have the TTS represent it in output
helps autisim spectrum
games with an avatar
enhance captioning
can be expressed in different ways, can just say the emotion, put an avatar with appropriate expression, etc.
--
EMMA represents semantics derived from less semanticful speech
always used for input, now starting to work on output
create a structured representation of intent that could be rendered in different forms
just starting
want to come up with a bunch of accessibility use cases
e.g., ways to simplify language
ls: interesting - adaptable language
dd: or symbols
ls: best way for dyslexics to learn is multimodal
dd: so redundant output
ls: various types of redundancies
can get really creative with it
dd: common to have systems that speak and display to screen, e.g., for noisy environments
cs: example of software to teach spelling
<scribe> meeting: PF FtF Day 2
ls: somewhat out of scope of cognitive but we´d want to list the benefits
cs: interested in authoring supports
two kinds: helping with a task at hand, and skill teaching
ls: that´s definitely out of scope, because @@
cs: things like spell checking though seems in scope
dd: work on making text easier to read
help author measure difficult of text and rewrite
and automatic text simplification
there are a whole bunch of readability forumulae
that quantify very specific things
heavily English-centric
can be gamed, hard-to-read text that passes the metrics
there has been validation of measurement via reading comprehensions tests
text simplification is early-stage
question how much text can be simplified
can change meaning
dd: separate modifiers into multiple sentences
ls: see that done with automatic translation
... added risk for cognitive
take translation, if you sort of know you can see that it got it wrong
but for cognitive simplification, can´t rely on user to pick that up
so might be more useful as support for authors to simplify, but needs to be checked author-side
one of the methodologies is alternative content
need metadata that allows people automatically find the right alternative content for their needs
these tools might help authors who prepare that alternative content do so better
cs: there is a lot of work out there on this
dd: intrigued by idea of automated simplification as a first step for authors
there are different language factors that affect readability
can be more difficult to measure
rare words, complexity of sentences, concrete vs abstract concepts, etc.
there is now work on combining natural language understanding with machine learning
machine learning might be able to create simplicity formulae from complex factors
has been done less than might expect
ls: has been done with Hebrew translation, reduce ambiguity re diacritic marks
<a little casual discussion scribe missed>
cs: Pearson may have some stuff
ls: Voice of America had some stuff
ls: sent call for volunteers yesterday to work on some of this
have some volunteers for sections (will be in wiki)
gap analysis now posted to http://www.w3.org/WAI/PF/cognitive-a11y-tf/wiki/Gap_Analysis
<covers how to add your sections>
ask for help with technology if needed, don´t let that impair content productivity
<scribe> ACTION: jason to find resources about classification of cognitive function [recorded in http://www.w3.org/2013/11/12-pf-minutes#action05]
<trackbot> Error finding 'jason'. You can review and register nicknames at <https://www.w3.org/WAI/PF/Group/track/users>.
ls: will add a brainstorming section to wiki
ls: would like to do user group research within three months
and rest of background research within 6 months
then do gap analysis in the three months after that
and build suggestions in the three months after that
so a year to complete the entire gap analysis
<tangent into possibilities of future FtF>
basically would like to have a lot of the background material ready for a FtF in roughly 6 months
<discussion of possible call times, will set up a scheduling poll next week>
Lisa summarizes the fields of the template (to be added to the wiki as a proposal).
Lisa edits the proposal live, with occasional clarificatory comments (of an editorial nature) from meeting participants.
<scribe> ACTION: LisaSeeman to post draft of template to the W3C wiki [recorded in http://www.w3.org/2013/11/12-pf-minutes#action06]
<trackbot> Error finding 'LisaSeeman'. You can review and register nicknames at <https://www.w3.org/WAI/PF/Group/track/users>.
<MichaelC> ACTION: Lisa to post draft of template to the W3C wiki [recorded in http://www.w3.org/2013/11/12-pf-minutes#action07]
<trackbot> Created ACTION-1308 - Post draft of template to the w3c wiki [on Lisa Seeman - due 2013-11-19].
<cyns> Persona: http://en.wikipedia.org/wiki/Persona_(user_experience)
<cyns> user story: http://en.wikipedia.org/wiki/User_story
<cyns> scenario: http://en.wikipedia.org/wiki/Scenario_(computing)
<cyns> use case: http://en.wikipedia.org/wiki/Use_case
There is discussion of the value of a persona (or user story), or a collection of representative ones, to capture the needs of people with specific cognitive disabilities.
The purpose is to describe cognitive function and the challenges encountered in using the Web.
It is decided to iterate the process of refining the format of the proposed documentation, starting from the templates and ideas discussed here.
Janina proposes to adjern the meeting.