Project:Home: Difference between revisions
(cwbn) |
|||
| Line 63: | Line 63: | ||
In order for the merger with the World News Wikibase to be complete, this should also be able to support querying domains by identified country. Also adding queryability by top-level domain would be useful as well. | In order for the merger with the World News Wikibase to be complete, this should also be able to support querying domains by identified country. Also adding queryability by top-level domain would be useful as well. | ||
{{cwbn}} | |||
Revision as of 01:00, 7 June 2023
Internet domain names, like wikibase.cloud, have peculiar properties. They are nominally identifiers to underlying resources, like nytimes.com and The New York Times. However, they are not persistent identifiers, as their value can change. Rather, publications and their owners are more like tenants over the domain name, and can decide to vacate the domain, eventually leading to a new owner.
This poses a significant challenge to modeling information sources. If a URL is identified as a reference in a Wikipedia article, and that URL is associated with a domain that has changed owners, the URL in question may or may not be a reliable source depending on when it was published.
As an interim solution, we have created the Internet Domains Wikibase, which creates items for individual domain names and then seeks to associate those with the publication or conceptual website they describe (via Wikidata item). Wikibase allows us to create nuanced statements through qualifiers, which is useful for describing the history of a domain name's ownership.
Items
Most items on this wiki are domain names. A domain name exists as a standalone entity apart from whatever companies or products are located at that domain name. Items with domain name items are assigned P1. values. Support items that are not domain names do not have P1 values.
Domains can then be associated with the publication or product contained at that domain with P2, with qualifiers as necessary.
Item:Q20419 (label -, description domain) is a honey trap for bad data. Somehow, data without domain names end up in imports, even though this is a wiki for domain names. They tend to use a single hyphen for the domain field. This item is useful in keeping track of this odd data, and should generally be left alone. (Changing the label or description affects its usability, as that is part of how data ends up on it.)
Properties
Create new ones as needed through Special:NewProperty.
URLs
- domain name (P1)
- formatter URL (P9)
- data record available at URL (P22)
Source Metadata
- title (P10)
- language used (P18)
Assessments
- appears on source listing (P7)
- assessed source reliability (P8)
- assessed source bias (P11)
- award received (P14)
- Misinfo.Me rating (P23)
Relations
- Wikidata item of associated publication or product (P2)
- exact equivalent Wikidata item (P5)
- exact equivalent Wikidata property (P6)
Time
- start time (P3)
- end time (P4)
- point in time (P16)
- inception (P20)
String representations
- category string (P12)
- publisher string (P13)
- qualifier string (P19)
- language string (P24)
- country string (P25)
- medium string (P26)
Other
- conferred by (P15)
- member of (P17)
- two-letter language code (P21)
Queries
- Domains and their item numbers
- Domains mapped to Wikidata items
- Domains appearing on source lists
- Domains appearing on both the Wikipedia-wide and vaccine safety-specific perennial sources lists
- Domains appearing on the most source lists
In order for the merger with the World News Wikibase to be complete, this should also be able to support querying domains by identified country. Also adding queryability by top-level domain would be useful as well.