Re: Suggested amendment to <a> tag

Hello,

the proposal implies some unsolved questions:

Search engines don't store a negative ranking. The resulting additional
data might be enormous. Basically all search engines rely on the basic
concept of "positive testimonials": the more often the link is present on
other websites the more interesting it is obviously.

For me it's rather impossible to say I agree or disagree with a paper of
someone else. Even though I agree with some statements I might disagree
with (some of) the conclusions or just some assumptions or a single
sentence. It might be even more difficult to agree or disagree with a
single webpage whole website.

I'm not sure how a instance of 'agree="false"' should be taken into account
and be displayed. How should it be taken into account for different
paths/resources on the same website? Or what does it mean for the judged
content if it is a image or list of facts?

For a review on products a review annotation schema as microformat/RDF
pattern is available.
This pattern should be easy adoptable for essays,  scientific papers and
virtually almost everything else.
The semantic essence of the agree attribute is IMHO to weak because a plain
disagreement doesn't tell enough about the value of the judged content. The
agreement or disagreement is better expressed in the text which is wrapped
in the link. In the given example it might be semantically unclear, if you
disagree with the websites content or the statement embodying the link.

At least two relevant dimensions are ignored in the proposal: trust and
quality.
Much more interesting for a search engine (and the user) than a (discrete)
level of agree of a (most probably) unknown entity would be the level of
trustworthiness and quality. The latter one might be seen as an abstract
metaview of agreement.

Both can be directly used to influence the ranking within a SERP.

Just my two cents

Jens Thorsten
Am 05.02.2013 11:48 schrieb "Richard Ferrers" <richard.ferrers@ands.org.au>:

> Dear html5spec team*
>
> Problem*: Google uses the <a> tag as a vote on the usefulness of the
> resource linked to.
> Not all links may be of the same or even positive value to the page
> creator.
>
> For instance:
> The <a href="http://www.ands.org.au">ANDS website</a> is an appalling
> example of web design.
>
> Google would register this link as data indicating the positive useful
> value of the target resource of the link. Yet the text around the link may
> suggest a negative value connection with the resource linked.
>
> *Solution*: I suggest a new attribute for the <a> tag to indicate whether
> you agree or disagree with the resource linked to.
> For example:
> The <a href="http://www.ands.org.au" agree=false>ANDS website</a> is an
> appalling example of web design.
>
> Google could see this 'agree' element and take it into account when
> ranking the page.
> Multiple inclusion of the link could indicate degree of
> negativity/positivity.
> Default: agree=true.
> Style: an alternate style could separate the status of the agree flag.
>
> For more detail a tag/s could be added to indicate the nature of the
> agreement/disagreement in the above examples eg tag=design.
>
> For your consideration.
> Richard
>
> --
> Richard Ferrers
> Research Data Analyst | Client Liaison Officer
>
> Australian National Data Service (ANDS)
> Physical Address: Level 6, Building F, Monash Caulfield VIC 3145 AUSTRALIA
> Postal Address: c/o Monash University, PO Box 197, Caulfield East, VIC
> 3145 AUSTRALIA
> T: +61 3 990 20569
> F: +61 3 990 20585
> M: 0422 368 061
> E: richard.ferrers@ands.org.au