W3C home > Mailing lists > Public > www-tag@w3.org > January 2012

Re: Is the TAG looking at "online filter bubbles"?

From: Jonathan A Rees <rees@mumble.net>
Date: Mon, 23 Jan 2012 20:53:10 -0500
Message-ID: <CAGnGFMLd2YPb65p6=awcP0rND9ENg7dk83hcgsRt24nXFMUK3A@mail.gmail.com>
To: Ian Jacobs <ij@w3.org>
Cc: www-tag@w3.org
I don't think the TAG is currently looking at this.

It certainly bothers me personally; I used to be able to say "google
for x y z and look at the second result" (or the URI equivalent) but
that no longer works. It probably interferes with the WPN idea
http://www.ibiblio.org/hhalpin/webpropernames/paper/, and so on. So I
would say the answer is yes, too much personalization threatens the
usefulness of the URIs.

The problem with webarch is that there is plausible deniability around
what the resource is, so the advice has no teeth - a hostile reading
of it makes it vacuous, since the resource can just be by fiat
whatever the web site coughs up. Google can (and does, implicitly)
just say: We mean for the URI to identify a user-relative resource.
That's just what it means; lump or leave it. The webarch advice could
be stronger - the URI should bring the recipient to a state that is
"as similar as possible" or "usefully similar" to the state seen by
the originating user, or something like that - carry the information
that A got to B - you sent the URI so that the recipient would see
what you see, right? - but it's hard to think how this could be made
rigorous enough to deal with the cases you raise, while still allowing
adequate flexibility to cover a variety of legitimate use scenarios.

I think a good comparison could be made to file: URIs, which are also
personalized references.

My 2 cents.


On Tue, Jan 17, 2012 at 11:28 PM, Ian Jacobs <ij@w3.org> wrote:
> Hello TAG,
> This TED talk crossed my desk:
>  http://www.ted.com/talks/eli_pariser_beware_online_filter_bubbles.html
> The question: does too much personalization undermine the universal information space?
> Is the TAG looking at this question? I was thinking about this as a trust issue:
>  * Webarch says [1]: "A URI owner SHOULD provide representations of the identified resource consistently and predictably."
>  * Predictability and consistency for me includes: "Other people will have a similar experience as mine for the same or similar URI"
>  * If they don't have a similar experience, then I would like to understand why. (For instance "Oh, you have a different account with difference settings, ok." or "Oh, you live in another country where the advertising is different.")
>  * If the reasons for a dissimilar experience are hidden from me, that erodes my trust in the service.
>  * It further erodes my trust if it becomes obvious that I have no way of understanding why I am having the tailored experience I am having, or that I may not be able to "turn it off."
> I was curious whether the TAG was discussing this.
> Ian
> [1] http://www.w3.org/TR/2004/REC-webarch-20041215/#URI-persistence
> --
> Ian Jacobs (ij@w3.org)    http://www.w3.org/People/Jacobs/
> Tel:                                      +1 718 260 9447
Received on Tuesday, 24 January 2012 01:53:39 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:33:13 UTC