Hey all,
I wanted to followup from the Zooniverse comments/thread a couple months
back, as part of Pharos's comments about crowd-sourcing depicts statements (
https://lists.wikimedia.org/pipermail/wikidata/2017-June/010795.html). I
met with Zooniverse staff during the week leading up to Wikimania, and was
able to get a rough outline of how they might be connected with our
ecosystem:
- They have a project builder system, that allows anyone to create a
project that uses their crowdsourcing structures, to do description and
identification of components of media files. The documentation is at:
https://www.zooniverse.org/lab
- They have an undocumented method for drawing media from external
repositories via a simple API -- so we could be calling on content
generated from sets via Commons or Wikidata.
- They would be interested in exploring if there is a good process for
connecting Commons as a source for their project builder, and would be
willing to provide some developer advisory support for either a volunteer
or partner institution to develop an appropriate link between the two apis.
I am talking with Pharos about potentially doing something with the Met set
on Commons, but if you think you have a set of media files already on
Commons that might be of interest to a citizen science-type crowdsourcing
project - please let me know. (Tools developed by the Wikidata community
might be more appropriate in the short term for the Met collection [1])
- Cool-less relevant note: Zooniverse is experimenting with using
Machine learning models to both prompt citizen science actions, and to sort
various sets of media in projects.
I am interested in exploring this relationship, because as Structured Data
on Commons gains the ability to store structured media file information in
the next few years, we have an increased ability to absorb simple
descriptive and other crowdsourcing on top of media files, and the
Zooniverse community provides access to a very wide group of
crowd-contribution interested communities. Moreover, if one of the value
statements for uploading to Commons, was a simple access point to
Zooniverse Crowdsourcing for either scientific purposes or enriching
structured descriptive metadata, we might have a lot more contributions of
both scientific and cultural heritage collections to Commons.[2]
If you have a set of media on Commons that you might be interested in
testing in the Zooniverse Project Builder, and have some developer
capacity, please let me know offlist. We don't need to move quickly on the
offer of consultation, but I would like to continue talking with Zooniverse
about how to create a better relationship.
Cheers,
Alex
--
Alex Stinson
GLAM-Wiki Strategist
Wikimedia Foundation
Twitter:@glamwiki/@sadads
[1] see the update from Gordibach:
https://meta.wikimedia.org/wiki/Grants_talk:IdeaLab/Wikidata_Paintbrush
[2] I am also keenly aware that any projects that pilot this kind of data
enrichment, will require a fair amount of Commons and/or Wikidata community
discussion about data quality, and its use.
Learn more about how the communities behind Wikipedia, Wikidata and other
Wikimedia projects partner with cultural heritage organizations:
http://glamwiki.org