Hoi,

The database design for OmegaWiki had a distinction between the concept and all the derivatives for them. The lexemes were all connected to the concept and independent of the spelling they are connected to the concept. Obviously this is language dependent.

So bumblebee is more complex than just "instance of" noun. It is an English noun. "Hommel" is connected as a Dutch noun for the same concept and "hommels" is the Dutch plural...

Obviously.

Thanks,

GerardM

On 13 September 2016 at 16:35, Daniel Kinzler <daniel.kinzler@wikimedia.de> wrote:

Am 13.09.2016 um 15:37 schrieb Gerard Meijssen:
> Hoi,
> You assume that it is not good to have lexicological information in our existing
> items. With Wiktionary support you bring such information on board. It would be
> really awkward when for every concept there has to be an item in two databases.

It will be two namespaces in the same project.

But we will not duplicate items. The proposed structure is not concept-centered
like Omegawiki is. It will be centered about lexemes, like Wiktionary is, but
with a higher level of granularity (a lexeme corresponds to one "morphological"
section on a Wiktionary page).

> Why is there this problem with lexicologival information and how will the
> current data be linked to the future "Wiktionary-data" information if there are
> to be two databases?

Because "bumblebee" <instance-of> "noun" conflicts with "bumblebee"
<subclass-of> "insect". They can't both be true for the same thing, because
nouns are not insects. One is true for the word, the other is true for the
concept. So they need to be treated separately.

--
Daniel Kinzler
Senior Software Developer

Wikimedia Deutschland
Gesellschaft zur Förderung Freien Wissens e.V.

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