2015-07-26 21:04 GMT+02:00 Andre Koopal andre@molens.org:
Hi,
I see a lot of discussion about how to get the data into wikidata. That is good and needs to be done, but erfgoedbot did a lot more.
Beside the harvesting (getting the list into the database), there were scripts using the data. At least for the dutch competition, people who uploaded a picture only specified the ID. Scripts used that id to find the data of that monument and add categories like “rijksmonument in <city>” instead of just “rijksmonument”. Also if it was known from the monument database it was a windmill (for example) it would be categorised as such.
Another important one is making lists of id’s where in the lists there is no picture but there is a picture uploaded. Volunteers would then look at that list and add a picture to the list. This worked quite well for the motivation, because people made a picture, uploaded it, and a couple of days later the picture showed up in the list.
If we think about the future, we will need to think about this side as well. The first one is needed to make the pictures being able to be found and not just a bunch of pictures. The first one is doable I guess, but I fear a bit about the second one. Because the lists are not regularly imported, you don’t know if there is a picture, so we need a way to detect this. I believe there are properties to document a picture, so the storing isn’t the problem I guess.
Regards,
André
Since the erfgoed database is going down, I guess erfgoedbotis will no longer be here anymore. Or maybe could it work with Wikidata ? Since Maarten don't want invest time anymore, I guess a new bot is needed.
Same thing goes for a lot of other tools like tools.wmflabs.org/wlm-maps
Meanwhile, there is Wikidata tools that can be esaly use for similar purposes (onsite there is the constraints reports and Autolist can generate list too).
Cdlt, ~nicolas