From: Felipe Ortega [mailto:firstname.lastname@example.org]
Sent: November 15, 2008 12:43 PM
To: Research into Wikimedia content and communities; Desilets, Alain
Subject: RE: [Wiki-research-l] "Regular contributor"
--- El vie, 14/11/08, Desilets, Alain <Alain.Desilets(a)nrc-cnrc.gc.ca>
De: Desilets, Alain
Asunto: RE: [Wiki-research-l] "Regular contributor"
Para: glimmer_phoenix(a)yahoo.es, "Research into Wikimedia content and
Fecha: viernes, 14 noviembre, 2008 2:32 Regarding this, I have had
heard different stories about contributors.
I seem to recall one study that concluded that, while 85% of the
**edits** are done by a small core of contributors, if you take a
random page and select a sentence from it, this sentence is more
likely to be the result of edits by contributors from the "long tail"
than core contributors. I forget the reference for that study though.
Does someone on this list have solid information about this? I think
it's a fairly crucial piece of information that we should have a
Hi, Alain. Yes, the study is by Aaron Schwartz. It was a base premise
in our last paper at HICSS 08, comparing his statement to the theory of
Jimmy Wales about the core of very active users.
Actually, both are right (more or less :) ). If you look at it from the
"per_user" perspective, the core can be identified very precisely.
But your question is focused on "per_article" statistics. It's logical
to expect so, since the distribution of distinct authors per article
follows a stepped power-law, and you have a lot of articles in the
larger editions. If you pick an article at random, chances are that you
will, most probably, pick one with few editors.
research-l-bounces(a)lists.wikimedia.org] On Behalf
Sent: November 13, 2008 5:33 PM
To: Research into Wikimedia content and communities
Subject: Re: [Wiki-research-l] "Regular
You have a very similar effect in larger Wikipedias.
In those ones,
there is no very active, "single
contributor, but a core of
very active users concentrating about 85% of the
number of edits
It seems that in these languages, though, there is a
in which new active users jump into the core to
substitute those who
eventually give up, for any reason. So, the
stable after a couple of years (aprox.) and the
encyclopedia is able to
--- El jue, 23/10/08, Gerard Meijssen
> De: Gerard Meijssen
> Asunto: Re: [Wiki-research-l] "Regular
> Para: "Research into Wikimedia content
> Fecha: jueves, 23 octubre, 2008 10:27 Hoi, I missed that this was
> the research mailing
list.. my fault.
> Consequently my answer was not appropriate.
this in mind, it is
> interesting to learn how the spread is in
particularly the smaller
> projects. In my opinion there must be a
amount of productive
> people in order to get to a community that
not have one person
> who is the "bus factor".
> Having someone who drives the bus is really
important. I wonder how
> you can point this person out. I think that
someone who is just
> editing is important but it is not all that
builds a community.
> On the Volapuk wikipedia Smeira was really
important. When he left, I
> understand that activity collapsed.
> 2008/10/22 phoebe ayers
> > 2008/10/21 Gerard Meijssen
> >> Hoi,
> >> When you divide people up in groups,
> single out the ones "most
> >> valuable", you in effect divide the
> community. Whatever you base your
> >> metrics on, there will be sound
arguments to deny
> the point of view. When it
> >> is about the number of edits, it is
clear to the
> pure encyclopedistas that
> >> most of the policy wonks have not
> is the "real" aim of the
> >> project.
> >> When you label groups of people, you
> and it is exactly the
> >> egalitarian aspect that makes the
> > But this isn't about labeling people for
> of time and saying that
> > this is how they are defined *on Wikipedia*
> it's about saying how do you
> > study people who regularly contribute to
> and as a part of that
> > how do you define the group that you are
> which is an important
> > question for any research study.
> > Given that it's impossible to study
> contributor to the project in
> > every study, and since many researchers are
> in why people who
> > spend a lot of time or effort working on
> so (and what exactly
> > it is they do), this is a very relevant
> this list.
> > --phoebe
> > Wiki-research-l mailing list
> > Wiki-research-l(a)lists.wikimedia.org
> Wiki-research-l mailing list
Wiki-research-l mailing list