[Foundation-l] Friendliness (was: Missing Wikipedians: An Essay)

[Andrew Gray]

On 25 February 2011 03:01, John Vandenberg <jayvdb at gmail.com> wrote:

> English Wikipedia is now sufficiently well known and culturally
> important, that 'we' no longer need to care about new contributors.
> Even if only 1% of new contributors work their way past the rejections
> and through our maze of rules, we will still have significant growth.

For content growth (which is broadly "ever bigger, ever better"), yes,
but not for community growth.

The absolute number of "active" community members on enwp peaked in
early 2007 and has been in a slow decline more or less steadily since
then; it's currently about two thirds what it was.

If we don't increase the rate at which we attract and retain new
contributors while we can, there's a real danger we could end up by
2020 or 2025 with a virtually moribund community - a small handful of
devoted vandal-fighters spending their days trying to keep millions of
pages clean and stable, and no influx of new users worth mentioning
because no-one has the time to cultivate it.

I'm not saying it's inevitable, but there's certainly no end of
examples of once-flourishing internet communities that have died that
sort of death by neglect, a spiral of spambots, vandals, and passing
once-off contributors leaving plaintive notes but with no real way to
restart a critical mass.

(Interestingly, the decline of editors is more or less proportional to
the overall editing rate - since the beginning of 2008, the ratio of
overall edits per month to highly active users has been about 10,000:1
- so in relative terms, the recent-changes "firehose" has been stable
for three years)

> We need systems which ensure that, on large projects, each newbie end
> up in contact with more than one established users who *care* about
> the specific topical area that the newbie is interested in.

There already is a relatively rough-and-ready system in place for
identifying and categorising new pages by project areas, using keyword
analysis:

http://en.wikipedia.org/wiki/User:AlexNewArtBot

producing daily reports like so:

http://en.wikipedia.org/wiki/User:AlexNewArtBot/IslamSearchResult

Building something which sends targeted invitation messages off the
back of that to new users is certainly plausible:

"Hi! You recently created [[Freedom and Justice Party (Egypt)]]. You
might be interested in the following projects working on these
topics..." with appropriate links and specific messages to, in this
case, the projects for Egypt/Africa/Politics/Law/Islam.

For people who don't create articles, you could have a bot look at the
first (say) ten or so article/talk edits of a new user, and then send
a list of suitable projects based on the way those pages were
categorised or project-tagged.

(I have no idea how easy this would be to implement...)

-- 
- Andrew Gray
  andrew.gray at dunelm.org.uk