Re: [Freebase-discuss] Propagation of bad sameAs statements

On Wed, Sep 8, 2010 at 10:59 AM, Philip Kendall
<philip-freebase@shadowmagic.org.uk> wrote:
> [ Crossposting. Apologies for the duplicate. ]
>
> ----- Forwarded message from Philip Kendall <philip-freebase@shadowmagic.org.uk> -----
>
> From: Philip Kendall <philip-freebase@shadowmagic.org.uk>
> To: freebase-discuss@freebase.com
> Subject: Re: [Freebase-discuss] Propagation of bad sameAs statements
> Date: Wed, 8 Sep 2010 15:55:31 +0100
>
> On Wed, Sep 08, 2010 at 10:42:45AM -0400, joel sachs wrote:
>>
>> And questions for freebase contributors: Are any of you running a script
>> that either a) loads in assertions from sameas.org, or b) deduces sameAs
>> relations from dbepedia redirection behaviour?
>
> Essentially, (b) - they're deduced from Wikipedia rather than dppedia,
> but it comes down to the same thing.
>
> I agree with you that it's the wrong thing to do - hopefully one of the
> Freebase Data Team will be along to explain why they do it.

It may be to allow any URL that refers (or referred) to a Wikipedia
page to be mechanically transformed into a valid Freebase URL, but
Wikipedia redirects are a mishmash of valid alternative names,
misspellings, and names of completely separate concepts which were
merged because they weren't big or significant enough to warrant their
own Wikipedia page.

I agree that it would be much better to have a single sameAs between
the concepts and to keep the information from the redirects as
alternate labels (if at all).

Speaking of DBpedia/Freebase sameAs links, the DBpedia side of things
shouldn't be using internal Freebase GUIDs.  They should either be
using the standard IDs or, preferably, the relatively new MIDs i.e.
one of the following:

  http://rdf.freebase.com/rdf/m.0hrk4
  http://rdf.freebase.com/rdf/en.invasive_species

As an aside, Freebase should also be using owl:sameAs to link these
alternate identities together.

Tom

Received on Wednesday, 8 September 2010 16:01:23 UTC