- From: Michael Pluke <Mike.Pluke@castle-consult.com>
- Date: Wed, 7 Oct 2015 15:58:05 +0000
- To: lisa.seeman <lisa.seeman@zoho.com>
- CC: public-cognitive-a11y-tf <public-cognitive-a11y-tf@w3.org>
- Message-ID: <A48C91EB13E45544B16FBC94C9298D8D3DD650@S11MAILD013N2.sh11.lan>
Hi Lisa As I am not very familiar with how all the pieces of the personalization approach work together it is not clear how the overriding would work. I guess that, somehow, the web author created ARIA importance settings need to have some importance values overridden by values stored in the personalization settings. However, this implies that in order for the “Sports” importance setting to be overridden, the user’s personalization settings need to include an entry that suggests that the user has an interest in sports clothing. The personalization settings would either need to include an alternative ARIA importance setting that exactly corresponds to the “Sports” item in the web page html file. Whereas it would be possible to include such override data for one frequently accessed website, I don’t see how this scales up to ensure that the user’s personal preference to always get information on sports clothes from ANY clothing website. I could see a possibility that a user’s personalization profile could include a preference that indicates the user’s interest in sports-related information. I could then see that there could be an intelligent user profile agent that understands the semantics of websites and, where it detects sports-related parts of the website, it ensures that all ARIA importance values associated with those parts of the site are given a “high” or even “critical” setting. An intelligent user agent is commonly assumed in sophisticated personalisation schemes like GPII and the ETSI personalisation work – but this is a long way removed from the simple approach that the use of ARIA importance suggests. So I still suggest that ARIA importance can be valuable for allowing some level of site simplification, but that it will be extremely difficult to extend this mechanism to personalizes a site to reflect an individual’s genuine needs and preferences (as a personalization scheme should). Best regards Mike From: lisa.seeman [mailto:lisa.seeman@zoho.com] Sent: 07 October 2015 12:09 To: Michael Pluke <Mike.Pluke@castle-consult.com> Cc: public-cognitive-a11y-tf <public-cognitive-a11y-tf@w3.org> Subject: RE: Key actions for this month Hi Mike I agree that importance is subjective to the user. Therefor it should be overidable by the personlization settings. All the best Lisa Seeman Athena ICT Accessibility Projects <http://accessibility.athena-ict.com> LinkedIn<http://il.linkedin.com/in/lisaseeman/>, Twitter<https://twitter.com/SeemanLisa> ---- On Wed, 07 Oct 2015 12:31:05 +0300 Michael Pluke<Mike.Pluke@castle-consult.com> wrote ---- As I am not fluent in ARIA or JSON I have significant difficulties assessing the proposal as the examples shown do not mean a lot to me. Taking things to a more conceptual level I would like to understand how “importance” is being handled. The “Importance” section says that identifying and differentiating between critical features, medium features and less important features of a web page “need to be defined as important from a user perspective”. I totally agree with this. But, if I understand things (and the nice clothes shopping demo) correctly, the display of “more” or “less” options is based on the “importance” of various aspects of the page contents based on the web author’s assessment of their “importance”. This assessment would be based on a statistical analysis of usage of a feature. If I understand the idea correctly the web author’s “importance” assessment would be based on an understanding on what percentage of the users use a feature. Although this statistical approach might be adequate for some features (i.e. probably a very small percentage of users will be interested in “Investors” feature, the real “importance” of a feature would be a highly personal experience. For example, at a clothes shopping site I would guess that 60% of the visitors might be women, and 40% men. So based on the draft, these would both have a “high” ARIA importance. But, at a personal level men’s clothes would be “high” for me but women’s clothes would be “low” (I am not brave or stupid enough to try to shop for clothes for my wife). The sample Demo script includes the “Sports” option in the initial and “Less options” states – because the web author has rated its importance as “high”. This is fine for me, but “Sports” is a “low” priority for my wife. So my question is whether the ARIA importance approach is really delivering personalization. To me it is providing a simplification mechanism that might be beneficial in some contexts. However, my fear is that it might also create site simplifications that are very unhelpful to many customers as it will remove options that are important to them. This type of “personalization” would be very unhelpful for anyone with atypical preferences. Many Autistic Spectrum users have their own intense interests and these may often be far removed from typical interests! Their unusual favourite interests would always be removed as soon as they attempts to simplify a complicated site. Best regards Mike From: lisa.seeman [mailto:lisa.seeman@zoho.com] Sent: 07 October 2015 07:31 To: public-cognitive-a11y-tf <public-cognitive-a11y-tf@w3.org<mailto:public-cognitive-a11y-tf@w3.org>> Subject: Key actions for this month Hi Folks Let us review the ARIA proposal in time for TPAC later this month. Please take a look and we will discuss it on Fridays call. See https://rawgit.com/w3c/coga/master/issue-papers/links-buttons.html Also WCAG has asked us to have a draft they can review by the end of the month. So also take a look at https://www.w3.org/WAI/PF/cognitive-a11y-tf/wiki/Proposal_for_WCAG All the best Lisa Seeman Athena ICT Accessibility Projects<http://accessibility.athena-ict.com> LinkedIn<http://il.linkedin.com/in/lisaseeman/>, Twitter<https://twitter.com/SeemanLisa>
Received on Wednesday, 7 October 2015 15:58:45 UTC