Hoi,
For the English Wikipedia there is a page where you find the most often
failed searches [1]. We have asked for the code for this software and we
received it. What we want is expand the functionality and use it for any
language.
When we do, we want to differentiate in failed searches that do not exist
in Wikidata either and the failed searches that exist in Wikidata.
As you may know searching Wikidata as well has been enabled on several
Wikipedias among them the Italian and the Polish Wikipedia. This allows us
to provide access to articles in other languages, it allows for finding
images in Commons because of a link to a Commons category, obviously access
to Wikidata and access in a more visual way care of the "Reasonator".
NB this functionality would add value to the en,wp as well because as you
may know 51% of the de.wp articles are linked to the en.wp...
What I am looking for is to have the developer who will modify the software
have access to the data. Magnus Manske is a well known and trusted
developer. He is the one that started MediaWiki. It is for him that I ask
access.
Our theory is that when we add Wikidata items, we will get more quickly at
the tipping point where search is actually useful (remember, there are 280+
Wikipedias). We expect that when we advertise what subjects do not have a
Wikipedia article but do have a Wikidata item, we will stimulate the
writing of articles that prove popular.
In effect we expect to engage in data driven user participation.
If you have any questions or suggestions I am happy to hear them. If
someone can get access to the data for Magnus please let me know.
Thanks,
Gerard