Also, Freebase users themselves who did daily, weekly work.... some where passing users, some tried harder, but made lots of erroneous entries (battling against our Experts at times).  We could probably provide a list of those sorta community blacklisted users who's data submissions should probably not be trusted.

+1 for looking at better maintained specific properties.
+1 for being cautious for some Freebase usernames and their entries.
+1 for trusting wholesale all of the Freebase Experts submissions.  We policed each other quite well.




On Fri, Sep 25, 2015 at 11:45 AM, Jason Douglas <jasondouglas@google.com> wrote:
> It would indeed be interesting to see which percentage of proposals are
> being approved (and stay in Wikidata after a while), and whether there
> is a pattern (100% approval on some type of fact that could then be
> merged more quickly; or very low approval on something else that would
> maybe better revisited for mapping errors or other systematic problems).

+1, I think that's your best bet. Specific properties were much better maintained than others -- identify those that meet the bar for wholesale import and leave the rest to the primary sources tool.

On Thu, Sep 24, 2015 at 4:03 PM Markus Krötzsch <markus@semantic-mediawiki.org> wrote:
On 24.09.2015 23:48, James Heald wrote:
> Has anybody actually done an assessment on Freebase and its reliability?
>
> Is it *really* too unreliable to import wholesale?

 From experience with the Primary Sources tool proposals, the quality is
mixed. Some things it proposes are really very valuable, but other
things are also just wrong. I added a few very useful facts and fitting
references based on the suggestions, but I also rejected others. Not
sure what the success rate is for the cases I looked at, but my feeling
is that some kind of "supervised import" approach is really needed when
considering the total amount of facts.

An issue is that it is often fairly hard to tell if a suggestion is true
or not (mainly in cases where no references are suggested to check). In
other cases, I am just not sure if a fact is correct for the property
used. For example, I recently ended up accepting "architect: Charles
Husband" for Lovell Telescope (Q555130), but to be honest I am not sure
that this is correct: he was the leading engineer contracted to design
the telescope, which seems different from an architect; no official web
site uses the word "architect" it seems; I could not find a better
property though, and it seemed "good enough" to accept it (as opposed to
the post code of the location of this structure, which apparently was
just wrong).

>
> Are there any stats/progress graphs as to how the actual import is in
> fact going?

It would indeed be interesting to see which percentage of proposals are
being approved (and stay in Wikidata after a while), and whether there
is a pattern (100% approval on some type of fact that could then be
merged more quickly; or very low approval on something else that would
maybe better revisited for mapping errors or other systematic problems).

Markus


>
>    -- James.
>
>
> On 24/09/2015 19:35, Lydia Pintscher wrote:
>> On Thu, Sep 24, 2015 at 8:31 PM, Tom Morris <tfmorris@gmail.com> wrote:
>>>> This is to add MusicBrainz to the primary source tool, not anything
>>>> else?
>>>
>>>
>>> It's apparently worse than that (which I hadn't realized until I
>>> re-read the
>>> transcript).  It sounds like it's just going to generate little warning
>>> icons for "bad" facts and not lead to the recording of any new facts
>>> at all.
>>>
>>> 17:22:33 <Lydia_WMDE> we'll also work on getting the extension
>>> deployed that
>>> will help with checking against 3rd party databases
>>> 17:23:33 <Lydia_WMDE> the result of constraint checks and checks
>>> against 3rd
>>> party databases will then be used to display little indicators next to a
>>> statement in case it is problematic
>>> 17:23:47 <Lydia_WMDE> i hope this way more people become aware of
>>> issues and
>>> can help fix them
>>> 17:24:35 <sjoerddebruin> Do you have any names of databases that are
>>> supported? :)
>>> 17:24:59 <Lydia_WMDE> sjoerddebruin: in the first version the german
>>> national library. it can be extended later
>>>
>>>
>>> I know Freebase is deemed to be nasty and unreliable, but is MusicBrainz
>>> considered trustworthy enough to import directly or will its facts
>>> need to
>>> be dripped through the primary source soda straw one at a time too?
>>
>> The primary sources tool and the extension that helps us check against
>> other databases are two independent things.
>> Imports from Musicbrainz have been happening since a very long time
>> already.
>>
>>
>> Cheers
>> Lydia
>>
>
>
> _______________________________________________
> Wikidata mailing list
> Wikidata@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikidata


_______________________________________________
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata

_______________________________________________
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata