Gerard Meijssen wrote:
Once we, as in the WMF, start paying for content there is no reasonable
argument to pay specific work and not pay for other specific work.
I am suggesting a limited experiment by the diverse chapters, not the WMF proper. I don't think it follows that success would mean there would be no reason to not pay for new content instead of citations for existing content.
Why should we pay for additional content in English and not pay for content in other languages?
CItation Hunt is already translated into six diverse languages, five of which are in our top 20, and it seems to work in RTL Hebrew.
Research is done that may lead to the use of Wikidata for citations.
I would love to see a link for the state of the art on that.
What is the status of https://meta.wikimedia.org/wiki/Research:Wikipedia_Knowledge_Graph_with_Deep...
and https://meta.wikimedia.org/wiki/Grants:IEG/StrepHit:_Wikidata_Statements_Val... ? There have been no updates on either at all this year, that I've been able to find, even though at least one of them is supposed to be producing monthly status reports. I'm not happy with the extent to which Wikidata has fallen into a meaningless soup of impenetrable data category numbers for its user interfaces. For a project being pushed as supporting translations, Wikidata is really difficult for direct use by humans speaking any natural language, but has been fantastic for search engines displaying summary information cutting into Foundation pageviews and fundraising.
We have a project called Wikiquote, why not invest attention into Wikiquote.
Because there is real demand and multiple quality-related fundamental needs for Wikipedia citations which can be easily automated with Citation Hunt, but all of the demand for Wikiquote expansion is fully addressed by Mediawiki as-is.
Really all the basic reasons why work on citations deserves additional funding is lacking. It does not explain what it will bring us anything that we cannot get in another way.
The number of {{citation needed}} tags is growing faster than new articles, and the rate at which they are addressed is so slow as to be negligible if you disregard WikiProject improvement drives, which occur less frequently than they used to.
As long as there is no obvious benefit, it would destroy what we are and how we do things for no obvious benefit.
As long as we don't measure the benefit, we have no way to know whether it's positive and will forever remain non-obvious.
On 23 April 2016 at 16:02, James Salsman jsalsman@gmail.com wrote:
Gerard Meijssen wrote: ...
I categorically oppose paying people for content. Enabling them to
create
content is different. Citations is content and its quality is relevant but only that.
Why categorically? We already pay hundreds of people for work in support
of
the projects, including reader-facing administrative and content far more prominent than citations. We encourage Wikipedian in Residence programs where third parties pay for all kinds of content development. The PR editing guidelines explicitly recognize that paid content happens anyway, we can't control it, but we can offer best practices. We support editing assigned as part of academic class requirements.
What reason is there to flatly rule out paying people to find citations before measuring the quality and cost/benefit ratio of doing so with a variety of both incentive payment models and managers?
How do people feel about a few of the larger the Chapters funding
pilots to have professional researchers do
https://tools.wmflabs.org/citationhunt/en
and a few other main languages?
It would be great to measure the quality of results of different
payment
incentive models and rates, but this is not something that the
Foundation
could do without some risk of breaching the DMCA safe harbor
provisions,
as
far as I can see. Even if I am technically wrong about that, the appearances would be that it's obvious exertion of what would be
positive
editorial control, which would still mean a greater likelihood of
lawsuits
by disgruntled BLP and corporate subjects who can't win in court but
can
waste everyone's money.
But I would rather have multiple measurements administered by
different
parties anyway, because there are likely to be large uncontrollable
sources
of noise.
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines New messages to: Wikimedia-l@lists.wikimedia.org