[Foundation-l] [EWW] Edit Wikipedia Week

Wed Nov 21 13:17:47 UTC 2007

Hi Chad 

Thanks for the response but really I was looking for a more thorough
methodology.  To say someone looks over it "to the best of their ability" is
not going to stand up in terms of assessment and benchmarking with regard to
authorities data or use of externally controlled authoritative data sources
as verification. 

What you need is a process model with usability attributes that have
designated measurable, unmeasurable or computable outcomes. Editor
credentials (IMHO) should be included within this process if you really want
something to be authoritative.

I can see that you have some computable outcomes which is a good start.

Best regards

Debbie

  _____  

From: Chad [mailto:innocentkiller at gmail.com] 
Sent: 21 November 2007 13:00
To: debbie at ictmarketing.co.uk; Wikimedia Foundation Mailing List
Subject: Re: [Foundation-l] [EWW] Edit Wikipedia Week

As a developer for Veropedia, I can personally tell you the checks we
perform on articles.

1 - No fairuse images. If an image is not freely licensed (GFDL, PD, CC, or
anything else applicable), we will not use it.
2 - No bad URLs. All external links must point to valid sites. 404s, 403s,
500s, all of these things are fixed or removed (whatever the case may be)
3 - Cleanup Categories/Templates - Any of the numerous cleanup categories
(external links, peacock terms, et cetera ad nauseum) as well as the
[citation needed] notices. 
4 - Disambiguations - All link need to point to a valid wiki article if
possible, disambiguations should be avoided.
5 - A variety of punctuation blacklists (refs on the wrong side of periods,
quotation marks inside other punctuation, et cetera) 
6 - We also check for slang, spelling and grammar problems outside of direct
quotes (our parser will actually notify the uploader on the more common
mistakes)
7 - Using too many of the same word will raise flags as well 
8 - Overly long or overly short sentences get flagged.

Also, we give our uploaders the Fog, Flesch and Flesch-Kincaid readability
indexes to help them in their editing.

Finally, there is the human element. Every article is (at least should, and
if they aren't, that person needs speaking to) read over by a real person
who judges (to the best of their ability) if an article is well written,
well sourced and well formatted. If they say yes, then they import that
revision to Veropedia where it remains frozen from editing (of course, new
versions can be imported over old ones, and we retain *our* upload history
as well as a link to Enwikis, a problem we are working on a feasible way to
fix right now). Hope this clears up some of the curiosity surrounding how we
decide articles are worthy of inclusion. 

-Chad H.

On Nov 21, 2007 7:45 AM, Debbie Garside <debbie at ictmarketing.co.uk> wrote:

Are there any documented general concepts, principles and requirements for
assessment and benchmarking articles within Wikipedia or veropedia?  Has a
usability/reliability/readability model been developed?  If so, can someone 
point me to a link?

As a standardizer, I would be interested to see the model for the "veropedia
test".

Best regards

Debbie

> -----Original Message-----
> From: foundation-l-bounces at lists.wikimedia.org
> [mailto:  <mailto:foundation-l-bounces at lists.wikimedia.org>
foundation-l-bounces at lists.wikimedia.org] On Behalf
> Of Dan Rosenthal

> Sent: 20 November 2007 23:48
> To: Wikimedia Foundation Mailing List
> Subject: Re: [Foundation-l] [EWW] Edit Wikipedia Week 
>

> Brian: before you continue talking, and noting that I still
> don't see you in #veropedia
>
> Veropedia gets it's articles by parsing a Wikipedia article, 
> generating a list of improvements (404s, disambigs, malformed
> templates, bad templates, readability indices etc.) and then
> the veropedian IMPROVES THE WIKIPEDIA ARTICLE until it passes
> the veropedia test, at which point it is uploaded. 
>
> There is no folding stuff back into WP. Waking up on the
> wrong side of the bed is no excuse for making a contentious
> statement on a topic you apparently know nothing about.
> Sorry, I tried to be nice about it in the last email, but 
> your response is plain childish.
>
> -Dan
> On Nov 20, 2007, at 6:33 PM, Brian McNeil wrote:
>
> > David Gerard wrote:
> >> On 20/11/2007, Waerth < waerth at asianet.co.th
<mailto:waerth at asianet.co.th> > wrote:
> >
> >>> Because we vannot do it all! Sometimes you need to branch off
> >>> specialistic projects to small groups of people. The Wikimedia
> >>> projects have grown so big that the head and the body 
> usually walk
> >>> in different directions and do different things. It is very
> >>> difficult to steer so many people. Like an earlier poster
> mentioned
> >>> .... consensus amongst such a huge body is impossible. That is 
> >>> easier reached amongst a smaller group of people. I hope more
> >>> initiatives like veropedia will arise!
> >
> >
> >> Open content: "Use our stuff. Please! (And give back your version 
> >> too.)"
> >
> > As the most succinct response on this I'll respond on this one.
> >
> > I get the message, the foundation can't do everything and
> the license 
> > allows
> > - nay - encourages projects like this. Good luck working out the
> > mechanics of the process of folding stuff back in to WP.
> >
> > Oh, and judging from some of the other posts today I'm not the only 
> > one that didn't actually fall out of the wrong side of bed but was
> > forcibly evicted before adequate sleep time had been acquired.
> >
> >
> > Brian McNeil
> > 
> >
> > _______________________________________________
> > foundation-l mailing list
> > foundation-l at lists.wikimedia.org
> > Unsubscribe: 
> http://lists.wikimedia.org/mailman/listinfo/foundation-l
>
>
> _______________________________________________ 
> foundation-l mailing list
> foundation-l at lists.wikimedia.org
> Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l
>
>
>

_______________________________________________
foundation-l mailing list
foundation-l at lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l