Re: Wiktionary parsing ; multiple languages

Le 2013-04-05 19:44, Sebastian Hellmann a écrit :
> Hi Mathieu,
>
> Am 05.04.2013 11:56, schrieb Mathieu Stumpf:
>> I added the dbpedia wiktionary entry on [1]. I wasn't aware of your 
>> effort, despite being really interesting in the wiktionary future. 
>> Could you please  read [1] and update it with your vision as a dbpedia 
>> contributor?
>>
>> [1] https://meta.wikimedia.org/wiki/Wiktionary_future
>>
>
> this page is interesting, but seems to be very idealistic. I am not
> sure, every language community agrees to use a common model. I also
> wonder if this is possible at all and whether there is an overlap. Do
> you think it makes sense to edit that page? Normally, there is a lot
> of talk and planning and nothing comes around in the end.

I would present that in an other way which would to say that this page 
try to adress the problem with long term perpectives, but with real 
concrete goals. Sure you can't reach the one solution that will make 
everybody happy, but making people talk together of their specifics 
issues and expectations from wikitionaries is a path which I think worth 
to be explored. To my mind, this should help us to have a better 
overview of various linguistic knowledge people are expecting to find in 
wiktionnaries, and how to improve the transmission of this knowledge 
between each chapters.

As it is said on the page, this is not a trivial problem, because it 
asks to gather a lot of linguistic expertise, as well as think about the 
UX we want to provide to end users and facilitate for third parties.

> Note that the good thing about Wiktionary is, that you can add
> information freely without adhering to a preset structure.

Yes and no. Sure if you don't count with the wikisyntax, there are no 
specific structure imposed to wiktionnaries chapters. But in practice, 
you know that they did adopted a more or less rigid structure, because 
that was relevant. But now we are in a situation where each chapter have 
its own idiom of templates, that not only make harder to automate 
cross-chapter information transmission, but also can make newcommers 
affraid. This is a really serious issue, I know that at least for the 
french chapter, we are losing wannabe contributor, because of heavy use 
we make of template. Don't get me wrong here, I'm not blaming the french 
wiktionary community, to my mind it's an upstream issue.

You know that having more editors is one of our community goals, don't 
you? Well, to have more editor, we have to make the participating 
leurning curve as small as possible. And that require a good UX. And 
that require a well thought end-user interface/API integration. I have 
no doubt it will be really difficult to integrate the Visual Editor into 
the french wiktionary for example, because articles there heavily relies 
on templates, and as far as I know, the Visual Editor doesn't provide 
(yet?) any tool to structure information further than 
section/bold/italic. But in the french wiktionary, even sections are 
created using templates!

>
> DBpedia is already implementing adapters to load data from WikiData.
> So Once WikiData is working for Wiktionary, we will have data from
> there and from the remaining Wikisyntax and merge them.DBpedia and
> WikiData have a loose cooperation for a joint task in a Google Summer
> of Code proposal.

Well, that's great, we need such a work to be done too. Thank you to do 
it.

_______________________________________________
Wiktionary-l mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l