I think separating the metadata makes sense because it cleans up the markup
language, and makes the metadata itself more easily extensible.
But, I do not think that the fact that the metadata is embedded in the pages
makes our trust/reputation computation more complex -- it would be fairly
easy to separate out the metadata at analysis stage, if we wanted to do
that.
On the negative side, separating the metadata requires new DB tables, etc
etc, making also the format of the dumps more complex (my guess).
In balance, though, it seems a good idea, but I am speaking here without
knowing the facts.
Luca
On Jan 30, 2008 1:41 PM, P. Birken <pbirken(a)gmail.com> wrote:
Currently, metadata is simply in the article body.
However, in theory,
metadata could be stored and edited in a separate place with its own
revision history, like the discussion is separate from the article. I
hope the idea is now more clear?
Philipp
2008/1/30, Luca de Alfaro <luca(a)dealfaro.org>rg>:
I am sure that, if it were of interest, one coudl
remove interwiki links
and
categories while analizing the pages... or am I
missing something?
I however would hesitate to remove the information from the revision
history: how to revert damage to the meta-information?
(Yes, better ways to compress the revision history are very much
needed,
for the working version, but this is another
issue)...
I hope my comments are on topic; I am not sure I understand the point.
Luca
On Jan 30, 2008 1:27 PM, P. Birken <pbirken(a)gmail.com> wrote:
Sorry for bringing this age old thread up again, but reading through
it again, I often thought: things would be easier, if Metadata were
not in the articles themselves, but separate. Then, the version
history would be trimmed dramatically and nobody could build up trust
(meaning both the real and the algorithmic sense) by category-edits
only. This would make life dramatically easier for the bots and easier
for users.
Is getting the metadata (category, interwikis, Personendaten on de)
out of the articles remotely feasible? Is the positive impact as good
as I think?
Best
Philipp
_______________________________________________
Wikiquality-l mailing list
Wikiquality-l(a)lists.wikimedia.org
http://lists.wikimedia.org/mailman/listinfo/wikiquality-l
_______________________________________________
Wikiquality-l mailing list
Wikiquality-l(a)lists.wikimedia.org
http://lists.wikimedia.org/mailman/listinfo/wikiquality-l
_______________________________________________
Wikiquality-l mailing list
Wikiquality-l(a)lists.wikimedia.org
http://lists.wikimedia.org/mailman/listinfo/wikiquality-l