User:Spinster/Pink ponies

Here's some stuff that I personally would find handy or useful in Wikidata.

If a solution already exists, or if you think I'm clueless, please tell me. In the first case, I will be exalted; in the second case I will cry in shame and will then proceed to improve my ways 😜

Robotic overlord helpers to add tedious referenced statements from external datasets

I try to dedicate some of my volunteer time to adding more referenced statements to items I care about. In many cases, I notice that I end up doing very boring, repetitive and brainless work. I tend to use one or a few reliable external datasets (some of my favorites are RKDartists (Q17299517), RKDimages (Q17299580) and Online Dictionary of Dutch Women (Q13135279)), read what's there, and copy some basic information from there (e.g. a person's gender, dob, dod, pob, pod, occupation(s), country of citizenship) with the same reference over and over (which, luckily, is made a bit less tedious with the DuplicateReferences gadget).

Here's an example of such a creation / addition process. Fleshing out a full item takes me 4 minutes (with practice, knowing the source website very well). But it's boooooring.

I know that Multichill does add some stuff by bot, and that Magnus has also figured out a way of adding some referenced dob&dod data via Mix'n'match. There's also the Primary sources tool (how it actually works, and whether new datasets can easily be added to it, is a bit obscure to me) and Soweego (which is all about identifiers, if I'm correct, so not really about adding those other statements?). I also assume that it requires quite some advanced/smart entity matching. But hey, if anyone has tips or ideas, I'd be grateful.

Something templatey

I sometimes teach Wikidata to newbies. Some of these newbies already have a lot of experience entering information in 'other' kinds of databases, but they still tend to get confused by Wikidata. One thing that confuses them most, is that it's quite unclear how you need to 'flesh out' a new item, and what information to add there.

This is what happens, and what (I think) would help:

• 
  What the Wikidata editing interface looks like, when creating a new item.
• 
  This confuses many newcomers. They don't know what to do here... What if, here, there would be a suggestion to choose if the current new item is one of several very commonly used ones (Q5 / human being one of them)?
• 
  A common mistake I see when newbies try to improve a new Wikidata item. They don't know how our statements and properties work, which ones exist, etc.
• 
  Why don't we do this? If an item 'knows' about itself that it's a Q5 (human), a lot of properties for potential statements could already be pre-suggested, making entry a lot more intuitive, especially for newcomers who don't know our property-value structure and our properties yet.
• 
  In fact, basic item structures are often defined by WikiProjects on Wikidata, and could be used as a source for this. This is an example of the basic item structure for a painting on Wikidata, as defined by volunteers in Wikidata's WikiProject Sum of all paintings.

I don't know how to call this 'pre-filled property/statement suggestor structure'. It reminds me of the templates on Commons, so I titled this section 'something templatey'.

Time

We can do time more accurately.

Faceted search

Wouldn't that be cool.

Integration of Mix'n'match in Wikidata proper

In Wikidata's search function, at item creation, and as part of items' identifier sections.