User talk:Hearvox

Jump to navigation Jump to search

About this board

Previous discussion was archived at User talk:Hearvox/Archive 1 on 2023-11-30.

Sdkb (talkcontribs)

I don't find the batch containing edits like this particularly helpful. All it indicates is that the newspaper was included in the database used by the project. Whereas I'd expect described by source (P1343) to statements to refer to a source that actually describes the item in some amount of detail. Could you please revert the batch until there is consensus for it?

Hearvox (talkcontribs)

The goal is to indicate credibility of news-sites in Wikidata. Inclusion in a vetted database of legitimate newspapers, like the News Desert project, is a credibility indicator. I consulted several active Wikidata editors who focus on news media and credibility. Since there's no property of inclusion in a database, they thought P1343 was the closest. If you know of a better prop for this purpose, please tell me.

Sdkb (talkcontribs)

What is your evidence that the database was meaningfully vetted?

Hearvox (talkcontribs)
Sdkb (talkcontribs)

That doesn't provide any information about how they compiled the database. And it seems that, to the extent there's a UNC database of newspapers, it'd be much better to include a statement providing the identifier of the entry in the database than one referencing a specific use case of the database, which includes no information other than the newspaper's name and county.

I am reverting the batch for now, without prejudice against restoring some reference to this project if consensus to do so in an appropriate form is reached.

Hearvox (talkcontribs)
Sdkb (talkcontribs)

That discussion has been archived, so it'd be better to start a new one.

Hearvox (talkcontribs)

These are all aliases for described by source: "entry", "mentioned in", "found in". All these alternate-labels are accurate descriptions of an entry in a database. Can you suggest a more appropriate property?

Reply to "News desert batch"
Elizium23 (talkcontribs)

Greetings, you appear to be adding Internet domain names as "English aliases" to various items. Can you please explain the rationale for this? It is not a common nor accepted practice. Please review your edits before we need to revert them for you, en masse.

Hearvox (talkcontribs)

I'm adding domain names to news-media as an alias. The domain of a newspaper is often synonymous with its name, and the URL is often the source of the article for a reader: "I read it in BostonGlobe.com" is the same as "I read it in the Boston Globe." Also, adding the domain as an alias makes the domain name searchable, which it wouldn't be otherwise. That is, a Wikidata search for "nytimes.com" only finds The New York Times if the domain was an alias. See demo of this: https://misinfocon.com/turning-wikimedia-into-a-news-site-credibility-tool-422dbf28fdec

Elizium23 (talkcontribs)

Well you're doing far more than that. You're adding it to human beings and other entities as well. There is a space for official website for every item we have; there is no need to clutter the "aliases" fields with these entries.

Hearvox (talkcontribs)

I should add many newspapers in Wikidata already have their domain name as an alias. I'm just making most news-media consistent with this practive, so both humans and machines can easily find a newspaper in WD by it's domain.

Hearvox (talkcontribs)

Additionally, domain names are (close to) unique, unlike the newspaper name itself (i.e., there are many named "Daily Times"). So its domain name connects a newspaper with their entries in external databases, thereby allowing us to import critical data into Wikidata: data related to source-reliability that Wikipedia now lacks (membership in a press association, for instance).

This post was hidden by Hearvox (history)
This post was hidden by Hearvox (history)
Reply to "Domain names"
There are no older topics