thoughts on community consensus in wikidata on wikicite-relevant stuff

[Arthur Smith]

I took the time to read through 2 previous (2013!) relevant RFC's on wikidata:

https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/References_and_sources

https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/Source_items_and_supporting_Wikipedia_sources

and I think the consensus there was quite clear and still stands - "Wikipedia sources should be supported" by creating items for each source (and for authors where necessary) - and create them in the main Q namespace, not in a separate namespace. So I think we are on solid ground proceeding with importing all wikipedia source data based on what the wikidata community has already decided.

Note that there were some specifics that may have an impact:

* for books an item for each book and (if necessary) each specific source edition used

* for scholarly, newspaper or magazine articles or similar entities an item for each article as well as for (and linked to) the journal/publisher

* for database entries, item is for database as a whole, reference should specify the exact entry

* but do NOT create items for individual web pages (reference URLs) - if part of a major website with an item that should be linked with "part of".

   Arthur

[James Hare]

Thank you for doing the research! This is by-and-large consistent with my plan for Librarybase. While we want journal articles on Wikidata, we can use Librarybase for more experimental kinds of data modeling like with websites.

--
Meta: https://meta.wikimedia.org/wiki/WikiCite
Twitter: https://twitter.com/wikicite
---
You received this message because you are subscribed to the Google Groups "wikicite-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wikicite-discu...@wikimedia.org.

[Marielle Volz]

Thanks.

I have a rather specific concern: they support adding in newspapers articles. However, there are large number of news articles now that blur the line between a blog post and newspaper and are not printed in any physical volume. And, in some cases, we are unable to distinguish between a news article and a blog post; currently in citoid there are currently blog posts that are currently incorrectly cast as newspaper articles:

https://phabricator.wikimedia.org/T161790

If we use citoid to add newspaper articles to wikidata, some will invariably be more of a blog post than a newspaper article. 

If this would be a problem, we might not want to add in 100% of "cite news" citations. 

Cheers,

Marielle

To unsubscribe from this group and stop receiving emails from it, send an email to wikicite-discuss+unsubscribe@wikimedia.org.

[Arthur Smith]

perhaps only add "news" posts if there is a known journal/newspaper/magazine that they are contained within (i.e. there is a wikidata item that they can be linked to as "part of")? I believe the thinking here is that if all there is to characterize a source is a URL (and a retrieval date, but that's not part of the source), then there's little advantage in having a distinct entity representing it - the URL is sufficient.

   Arthur

[Charles Matthews]

On Thursday, May 25, 2017 at 10:04:37 PM UTC+1, Arthur Smith wrote:

perhaps only add "news" posts if there is a known journal/newspaper/magazine that they are contained within (i.e. there is a wikidata item that they can be linked to as "part of")?

 

Or use "published in" (P1433) where appropriate. It is certainly useful generally in source criticism to understand when a part of a publication is subject to editorial control (or not). 

Charles