distinguishing native contributors from helpful strangers

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

distinguishing native contributors from helpful strangers

Amir E. Aharoni
Hi,

There is a phenomenon in Wikipedias in smaller languages: There activity
level of people who actually know the language of the wiki and make
meaningful text contributions is relatively low, and the activity of people
from other wikis who make various technical edits that don't require the
knowledge of the language is relatively high.

I call the latter group "helpful strangers". They can do things such as
fixing categories, fixing invalid wiki syntax, editing templates, adding
images, etc.—things that don't require knowing the language well, and can
be achieved by copying and pasting, by guessing things from interlanguage
links, or by writing language-neutral things, such as numbers or filenames.

Now, I've written "relatively low" and "relatively high", but these are
just my anecdotal impressions. Has anyone thought of a way to quantify this
more precisely?

--
Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי
http://aharoni.wordpress.com
‪“We're living in pieces,
I want to live in peace.” – T. Moore‬
_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Reply | Threaded
Open this post in threaded view
|

Re: distinguishing native contributors from helpful strangers

Marc Miquel
Hi Amir,

This is an interesting idea. I haven't found a way to detect whether an
editor is native or not. My approach to multingual editing is through the
concept of having a primary-non primary Wikipedias. Your primary Wikipedia
is the one where you have made more edits to (and you are a primary editor
there). In the rest of Wikipedias where you have at least an edit, you are
a non-primary editor.

I'm currently creating a database in which for every Wikipedia I have a
table with a column specifying whether an editor is primary from this
language or non-primary, another one with the primary language, another one
with how many other languages they interacted with and a final one with the
total number of edits in all languages.

An editor behaves quite differently when he is primary or non-primary in
terms of social interactions, topical diversity, etc. To me, this is
interesting because it allows me to detect when an editor "exports" content
(edits content about their local area, usually politics-related content, in
other languages).

Assessing the impact of these "technical helpful editors" may not be easy
as we'd need to examine the characteristics of the edits. However,
quantifying the extent of edits made by non-primary editors is doable.
Would that help you?

Best,

Marc Miquel

Missatge de Amir E. Aharoni <[hidden email]> del dia dc., 5
de juny 2019 a les 10:54:

> Hi,
>
> There is a phenomenon in Wikipedias in smaller languages: There activity
> level of people who actually know the language of the wiki and make
> meaningful text contributions is relatively low, and the activity of people
> from other wikis who make various technical edits that don't require the
> knowledge of the language is relatively high.
>
> I call the latter group "helpful strangers". They can do things such as
> fixing categories, fixing invalid wiki syntax, editing templates, adding
> images, etc.—things that don't require knowing the language well, and can
> be achieved by copying and pasting, by guessing things from interlanguage
> links, or by writing language-neutral things, such as numbers or filenames.
>
> Now, I've written "relatively low" and "relatively high", but these are
> just my anecdotal impressions. Has anyone thought of a way to quantify this
> more precisely?
>
> --
> Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי
> http://aharoni.wordpress.com
> ‪“We're living in pieces,
> I want to live in peace.” – T. Moore‬
> _______________________________________________
> Wiki-research-l mailing list
> [hidden email]
> https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
>
_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Reply | Threaded
Open this post in threaded view
|

Re: distinguishing native contributors from helpful strangers

Andy Mabbett-2
In reply to this post by Amir E. Aharoni
On Wed, 5 Jun 2019 at 09:42, Amir E. Aharoni
<[hidden email]> wrote:

> There is a phenomenon in Wikipedias in smaller languages: There activity
> level of people who actually know the language of the wiki and make
> meaningful text contributions is relatively low, and the activity of people
> from other wikis who make various technical edits that don't require the
> knowledge of the language is relatively high.

> Now, I've written "relatively low" and "relatively high", but these are
> just my anecdotal impressions. Has anyone thought of a way to quantify this
> more precisely?

It won't answer the question fully, but you can narrow down the
results by looking at babel templates to see which languages they
self-rate as being proficient in, or otherwise, on their home
project(s).

I try to act as a "helpful stranger" on non-English projects, for
instance by adding images and {{Authority control}} templates. This is
usually well received, but there are a couple of projects where the
former at least is apparently not welcome, and I've recently been
blocked (with no warning; my talk page ink is still red), with no talk
page or email access, on Lithuanian Wikipedia. In 2015 I was accused
of "vandalism" and "trolling" there.

Happy to discuss my experiences - good and bad - off-list, if that
will help your research.

--
Andy Mabbett
@pigsonthewing
http://pigsonthewing.org.uk

_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Reply | Threaded
Open this post in threaded view
|

Re: distinguishing native contributors from helpful strangers

Kiril Simeonovski
Hi all,

I think this is an excellent research topic that might give us helpful
insights on how Wikipedias can benefit from the support provided by
non-speakers. Discerning the namespaces where this support ends and whether
it was made by humans or bots may also give highly useful information. My
observations so far regarding any support by the so-called "helpful
strangers" can be summarised in the following conclusions:

* The larger the community size of a Wikipedia, the higher rules-lawyering
applied to the "helpful strangers". This means that:
** Very small Wikipedias (less than 25 active contributors) do not have a
strict set of rules nor a native-speaking contributors to watch and every
kind of support is welcome (mostly in the form of bot-generated articles
and automatic translation of templates).
** Small Wikipedias (from 25-100 active contributors) do have some set of
rules and some native-speaking contributors but most kinds of support are
still welcome.
** Medium-sized Wikipedias (from 100-1,000 active contributors) do have a
clear set of rules and a native-speaking community to take care of
everything; the room for support is limited to human editing that abides
some rules and sometimes community permission is required (mostly comes in
the form of categorisation an correction of templates, while bot-generated
stuff is mostly done by native speakers with a bot flag required for
strangers).
** Large Wikipedias (over 1,000 active contributors) do have rules about
things that could have not been imagined and native-speaking community that
easily manages the fields where the strangers could help in, making them
not attractive for non-native speakers to come in and help.

Another dimension could be a research on the block log of the "helpful
strangers" that might explain how these contributors are accepted by the
communities they are helping to.

Best regards,
Kiril

On Wed, Jun 5, 2019 at 13:24 Andy Mabbett <[hidden email]> wrote:

> On Wed, 5 Jun 2019 at 09:42, Amir E. Aharoni
> <[hidden email]> wrote:
>
> > There is a phenomenon in Wikipedias in smaller languages: There activity
> > level of people who actually know the language of the wiki and make
> > meaningful text contributions is relatively low, and the activity of
> people
> > from other wikis who make various technical edits that don't require the
> > knowledge of the language is relatively high.
>
> > Now, I've written "relatively low" and "relatively high", but these are
> > just my anecdotal impressions. Has anyone thought of a way to quantify
> this
> > more precisely?
>
> It won't answer the question fully, but you can narrow down the
> results by looking at babel templates to see which languages they
> self-rate as being proficient in, or otherwise, on their home
> project(s).
>
> I try to act as a "helpful stranger" on non-English projects, for
> instance by adding images and {{Authority control}} templates. This is
> usually well received, but there are a couple of projects where the
> former at least is apparently not welcome, and I've recently been
> blocked (with no warning; my talk page ink is still red), with no talk
> page or email access, on Lithuanian Wikipedia. In 2015 I was accused
> of "vandalism" and "trolling" there.
>
> Happy to discuss my experiences - good and bad - off-list, if that
> will help your research.
>
> --
> Andy Mabbett
> @pigsonthewing
> http://pigsonthewing.org.uk
>
> _______________________________________________
> Wiki-research-l mailing list
> [hidden email]
> https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
>
_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Reply | Threaded
Open this post in threaded view
|

Re: distinguishing native contributors from helpful strangers

Ziko van Dijk-3
Hello Amir,

Interesting, I called this phenomenon "foreigh helpers" nearly 10 years ago:
https://de.wikipedia.org/wiki/Benutzer:Ziko/Handbuch-Allgemeines
These people do not speak the language of the language version -
otherwise they would be simply a 'normal' part of the community. But
they help out with skills, maybe especially those that are not
existing in the small community.
I have considered also a foreign helper who is supporting translation
tools or platforms, like GerardM with providing scripts or people who
support translatewiki. But that is usually an activity outside the
wiki.

Kind regards.
Ziko




Am Mi., 5. Juni 2019 um 16:04 Uhr schrieb Kiril Simeonovski
<[hidden email]>:

>
> Hi all,
>
> I think this is an excellent research topic that might give us helpful
> insights on how Wikipedias can benefit from the support provided by
> non-speakers. Discerning the namespaces where this support ends and whether
> it was made by humans or bots may also give highly useful information. My
> observations so far regarding any support by the so-called "helpful
> strangers" can be summarised in the following conclusions:
>
> * The larger the community size of a Wikipedia, the higher rules-lawyering
> applied to the "helpful strangers". This means that:
> ** Very small Wikipedias (less than 25 active contributors) do not have a
> strict set of rules nor a native-speaking contributors to watch and every
> kind of support is welcome (mostly in the form of bot-generated articles
> and automatic translation of templates).
> ** Small Wikipedias (from 25-100 active contributors) do have some set of
> rules and some native-speaking contributors but most kinds of support are
> still welcome.
> ** Medium-sized Wikipedias (from 100-1,000 active contributors) do have a
> clear set of rules and a native-speaking community to take care of
> everything; the room for support is limited to human editing that abides
> some rules and sometimes community permission is required (mostly comes in
> the form of categorisation an correction of templates, while bot-generated
> stuff is mostly done by native speakers with a bot flag required for
> strangers).
> ** Large Wikipedias (over 1,000 active contributors) do have rules about
> things that could have not been imagined and native-speaking community that
> easily manages the fields where the strangers could help in, making them
> not attractive for non-native speakers to come in and help.
>
> Another dimension could be a research on the block log of the "helpful
> strangers" that might explain how these contributors are accepted by the
> communities they are helping to.
>
> Best regards,
> Kiril
>
> On Wed, Jun 5, 2019 at 13:24 Andy Mabbett <[hidden email]> wrote:
>
> > On Wed, 5 Jun 2019 at 09:42, Amir E. Aharoni
> > <[hidden email]> wrote:
> >
> > > There is a phenomenon in Wikipedias in smaller languages: There activity
> > > level of people who actually know the language of the wiki and make
> > > meaningful text contributions is relatively low, and the activity of
> > people
> > > from other wikis who make various technical edits that don't require the
> > > knowledge of the language is relatively high.
> >
> > > Now, I've written "relatively low" and "relatively high", but these are
> > > just my anecdotal impressions. Has anyone thought of a way to quantify
> > this
> > > more precisely?
> >
> > It won't answer the question fully, but you can narrow down the
> > results by looking at babel templates to see which languages they
> > self-rate as being proficient in, or otherwise, on their home
> > project(s).
> >
> > I try to act as a "helpful stranger" on non-English projects, for
> > instance by adding images and {{Authority control}} templates. This is
> > usually well received, but there are a couple of projects where the
> > former at least is apparently not welcome, and I've recently been
> > blocked (with no warning; my talk page ink is still red), with no talk
> > page or email access, on Lithuanian Wikipedia. In 2015 I was accused
> > of "vandalism" and "trolling" there.
> >
> > Happy to discuss my experiences - good and bad - off-list, if that
> > will help your research.
> >
> > --
> > Andy Mabbett
> > @pigsonthewing
> > http://pigsonthewing.org.uk
> >
> > _______________________________________________
> > Wiki-research-l mailing list
> > [hidden email]
> > https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
> >
> _______________________________________________
> Wiki-research-l mailing list
> [hidden email]
> https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Reply | Threaded
Open this post in threaded view
|

Re: distinguishing native contributors from helpful strangers

Aaron Halfaker-3
Looks like these kind of methods would be useful:
https://www.aaai.org/ocs/index.php/ICWSM/ICWSM16/paper/viewPDFInterstitial/13077/12764

I've been looking to implement edit type modeling in ORES.  You could use
something like this to build a profile on each editor by what types of work
they generally do on articles.  Right now, this has been deprioritized, but
I've got some interested from Wikimedia Product ("Audiences"), so we might
prioritize it soon.  I'd love to pull in a contractor to do the work to
bring it to ORES.

On Tue, Jun 11, 2019 at 9:15 AM Ziko van Dijk <[hidden email]> wrote:

> Hello Amir,
>
> Interesting, I called this phenomenon "foreigh helpers" nearly 10 years
> ago:
> https://de.wikipedia.org/wiki/Benutzer:Ziko/Handbuch-Allgemeines
> These people do not speak the language of the language version -
> otherwise they would be simply a 'normal' part of the community. But
> they help out with skills, maybe especially those that are not
> existing in the small community.
> I have considered also a foreign helper who is supporting translation
> tools or platforms, like GerardM with providing scripts or people who
> support translatewiki. But that is usually an activity outside the
> wiki.
>
> Kind regards.
> Ziko
>
>
>
>
> Am Mi., 5. Juni 2019 um 16:04 Uhr schrieb Kiril Simeonovski
> <[hidden email]>:
> >
> > Hi all,
> >
> > I think this is an excellent research topic that might give us helpful
> > insights on how Wikipedias can benefit from the support provided by
> > non-speakers. Discerning the namespaces where this support ends and
> whether
> > it was made by humans or bots may also give highly useful information. My
> > observations so far regarding any support by the so-called "helpful
> > strangers" can be summarised in the following conclusions:
> >
> > * The larger the community size of a Wikipedia, the higher
> rules-lawyering
> > applied to the "helpful strangers". This means that:
> > ** Very small Wikipedias (less than 25 active contributors) do not have a
> > strict set of rules nor a native-speaking contributors to watch and every
> > kind of support is welcome (mostly in the form of bot-generated articles
> > and automatic translation of templates).
> > ** Small Wikipedias (from 25-100 active contributors) do have some set of
> > rules and some native-speaking contributors but most kinds of support are
> > still welcome.
> > ** Medium-sized Wikipedias (from 100-1,000 active contributors) do have a
> > clear set of rules and a native-speaking community to take care of
> > everything; the room for support is limited to human editing that abides
> > some rules and sometimes community permission is required (mostly comes
> in
> > the form of categorisation an correction of templates, while
> bot-generated
> > stuff is mostly done by native speakers with a bot flag required for
> > strangers).
> > ** Large Wikipedias (over 1,000 active contributors) do have rules about
> > things that could have not been imagined and native-speaking community
> that
> > easily manages the fields where the strangers could help in, making them
> > not attractive for non-native speakers to come in and help.
> >
> > Another dimension could be a research on the block log of the "helpful
> > strangers" that might explain how these contributors are accepted by the
> > communities they are helping to.
> >
> > Best regards,
> > Kiril
> >
> > On Wed, Jun 5, 2019 at 13:24 Andy Mabbett <[hidden email]>
> wrote:
> >
> > > On Wed, 5 Jun 2019 at 09:42, Amir E. Aharoni
> > > <[hidden email]> wrote:
> > >
> > > > There is a phenomenon in Wikipedias in smaller languages: There
> activity
> > > > level of people who actually know the language of the wiki and make
> > > > meaningful text contributions is relatively low, and the activity of
> > > people
> > > > from other wikis who make various technical edits that don't require
> the
> > > > knowledge of the language is relatively high.
> > >
> > > > Now, I've written "relatively low" and "relatively high", but these
> are
> > > > just my anecdotal impressions. Has anyone thought of a way to
> quantify
> > > this
> > > > more precisely?
> > >
> > > It won't answer the question fully, but you can narrow down the
> > > results by looking at babel templates to see which languages they
> > > self-rate as being proficient in, or otherwise, on their home
> > > project(s).
> > >
> > > I try to act as a "helpful stranger" on non-English projects, for
> > > instance by adding images and {{Authority control}} templates. This is
> > > usually well received, but there are a couple of projects where the
> > > former at least is apparently not welcome, and I've recently been
> > > blocked (with no warning; my talk page ink is still red), with no talk
> > > page or email access, on Lithuanian Wikipedia. In 2015 I was accused
> > > of "vandalism" and "trolling" there.
> > >
> > > Happy to discuss my experiences - good and bad - off-list, if that
> > > will help your research.
> > >
> > > --
> > > Andy Mabbett
> > > @pigsonthewing
> > > http://pigsonthewing.org.uk
> > >
> > > _______________________________________________
> > > Wiki-research-l mailing list
> > > [hidden email]
> > > https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
> > >
> > _______________________________________________
> > Wiki-research-l mailing list
> > [hidden email]
> > https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
>
> _______________________________________________
> Wiki-research-l mailing list
> [hidden email]
> https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
>


--

Aaron Halfaker

Principal Research Scientist

Head of the Scoring Platform team
Wikimedia Foundation
_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l