Jan, Stas, Markus, all,
the run is now complete. I copied the ~13K queries, and the wikis/pages
they are used on, to
(128KB, bzip2)
The queries can be either WDQ or SPARQL; I guess SPARQL ones will start
with SELECT ;-)
Note: I do auto-replace "simple" (by heuristic) WDQ with SPARQL via the
wdq2sparql tool, so there is some bias there.
Cheers,
Magnus
On Fri, Oct 21, 2016 at 11:23 AM Jan Dittrich <jan.dittrich(a)wikimedia.de>
wrote:
That’s great!
I wonder how Stas and Markus want to classify the data, but if our
interest goes in the same direction, we could pool brains and resources to
find a classification or at least some nice ggplot vizs :-)
Jan
2016-10-20 14:42 GMT+02:00 Magnus Manske <magnusmanske(a)googlemail.com>om>:
Jan, my Listeria bot currently manages ~13K lists on various WMF wikis. I
just changed my update bot to collect all the queries that are used for
those lists. It will take a day or so to collect them, then I'll forward
them to you. Should be a good data set to estimate baseline demand, in
terms of what complexity is required.
On Thu, Oct 20, 2016 at 1:27 PM Jan Dittrich <jan.dittrich(a)wikimedia.de>
wrote:
Hi Magnus,
Hi List,
Thanks for your assessment!
Since "proper" wikidata lists would be
powered by SPARQL, which is much
more expressive, it will be difficult to wrap it
into a flowchart-like
interface, without overloading it with options, which in turn makes it
harder to use.
Yes, my goal would be not graphically wrap the whole SPARQL; like you, I
noticed it would be too overloaded. In any new interface, I would also only
support a subset – but a useful one (specifically, useful for querying to
generate Wikipedia lists, at least for now).
For example, one of my current drafts allows querying in this very
limited way: *Prop:Value AND (Prop:Value OR Prop:Value) AND*… – so
- only one Type (Query:Value, which also should work for Qualifiers,
though),
- one first level Operator (AND),
- one second level Operator (OR).
So I aim for:
1. Finding out which instructions are most useful in practice, to see
what might be included in such a subset (possibly like: "In almost all
queries, I use the ____-Operator, but I never understood what _____ was
for".)
and, which is why I asked about wqd,
2. What did work well about the using the graphical interface of wdq?
(possible things: How scary/friendly is it compared to SPARQL? How hard is
finding the needed instructions compared to SPARQL etc.)
Kind Regards,
Jan
2016-10-20 13:55 GMT+02:00 Magnus Manske <magnusmanske(a)googlemail.com>om>:
Hi Jan,
as the author of wdq and its query builder, I recommend against using it
as a model.
The wdq query builder does work, to some degree, because the instruction
set of wdq is very limited, and predictable. Since "proper" wikidata lists
would be powered by SPARQL, which is much more expressive, it will be
difficult to wrap it into a flowchart-like interface, without overloading
it with options, which in turn makes it harder to use.
Personally, I would let people just paste SPARQL into the list
definitions, and let them construct and test the queries elsewhere.
query.wikidata.org already has nice features, such as CTRL-space
replacing of free text by Wikidata entities, searchable examples, etc.
Combine that with a place where people can get help with query building,
and IMHO it would serve the community well.
There also was a natural-language-to-SPARQL tool somewhere, but I can't
find the URL...
This is not to say that we shouldn't have, at some point down the road, a
query-building interface that is tailored to our needs!
My two Eurocent,
Magnus
On Thu, Oct 20, 2016 at 12:30 PM Jan Dittrich <jan.dittrich(a)wikimedia.de>
wrote:
Hello everyone,
I continue to work on creating an interface that allows easy querying of
Wikidata for generating lists for wikipedia (and, possibly, beyond).
An existing, interface based query builder is
http://wdq.wmflabs.org/
If you use (or have used it) it would be interested in hearing about the
advantages or disadvantages you perceived in regards to the functions of
wqd and their interface.
Kind Regards,
Jan
PS: In my experience, examples ("I tried to…") or context ("because I
wanted to…") can greatly easy understanding, so if giving them makes sense
for you, don't hesitate to include them.
--
Jan Dittrich
UX Design/ User Research
Wikimedia Deutschland e.V. | Tempelhofer Ufer 23-24 | 10963 Berlin
Phone: +49 (0)30 219 158 26-0
http://wikimedia.de
Imagine a world, in which every single human being can freely share in the
sum of all knowledge. That‘s our commitment.
Wikimedia Deutschland - Gesellschaft zur Förderung Freien Wissens e. V.
Eingetragen im Vereinsregister des Amtsgerichts Berlin-Charlottenburg unter
der Nummer 23855 B. Als gemeinnützig anerkannt durch das Finanzamt für
Körperschaften I Berlin, Steuernummer 27/029/42207.
_______________________________________________
Wikidata mailing list
Wikidata(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
_______________________________________________
Wikidata mailing list
Wikidata(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
--
Jan Dittrich
UX Design/ User Research
Wikimedia Deutschland e.V. | Tempelhofer Ufer 23-24 | 10963 Berlin
Phone: +49 (0)30 219 158 26-0
http://wikimedia.de
Imagine a world, in which every single human being can freely share in the
sum of all knowledge. That‘s our commitment.
Wikimedia Deutschland - Gesellschaft zur Förderung Freien Wissens e. V.
Eingetragen im Vereinsregister des Amtsgerichts Berlin-Charlottenburg unter
der Nummer 23855 B. Als gemeinnützig anerkannt durch das Finanzamt für
Körperschaften I Berlin, Steuernummer 27/029/42207.
_______________________________________________
Wikidata mailing list
Wikidata(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
_______________________________________________
Wikidata mailing list
Wikidata(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
--
Jan Dittrich
UX Design/ User Research
Wikimedia Deutschland e.V. | Tempelhofer Ufer 23-24 | 10963 Berlin
Phone: +49 (0)30 219 158 26-0
http://wikimedia.de
Imagine a world, in which every single human being can freely share in the
sum of all knowledge. That‘s our commitment.
Wikimedia Deutschland - Gesellschaft zur Förderung Freien Wissens e. V.
Eingetragen im Vereinsregister des Amtsgerichts Berlin-Charlottenburg unter
der Nummer 23855 B. Als gemeinnützig anerkannt durch das Finanzamt für
Körperschaften I Berlin, Steuernummer 27/029/42207.
_______________________________________________
Wikidata mailing list
Wikidata(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata