Skip to content

Latest commit

 

History

History
52 lines (37 loc) · 5.05 KB

README.md

File metadata and controls

52 lines (37 loc) · 5.05 KB

URIs

This is a place to align URIs among different projects for persons, places, works, manuscripts, and potentially other entities related to the historical Middle East.

How to use aligned URIs

Here you'll find folders for each entity containing simple TSV tables with at least three columns. Each row represents a relationship between the URIs in each of the columns. The columns are not ordered: which URI appears in which column does not matter. Each URI may appear multiple times in the spreadsheet when it has relationships to additional URIs.

You can import this into the spreadsheet editor or data format of your choice and use lookup, filter, or similar functions to search for one URI and get related URIs from other domains (projects). Note: When searching for URIs, make sure to take into account both http:// and https:// versions as well as with or without www . In regex, this can be done with https?://(www\.)?.

This alignment assumes only that the URIs are related (formally, dc:relation can be asserted). It does not always mean that the URIs refer to exactly the same entity. For example, different projects might use different strategies to identify pseudonymous authors of works (e.g., "Pseudo-Chrysostom" vs. "Author of Homily X") or to connect historical places with geographical locations. These differing strategies mean that related URIs might refer to somewhat differently conceived entities. Formally speaking, owl:sameAs cannot be asserted. This is also the case because the list may contain URIs that are used to identify and disambiguate entities (such as a Wikipedia page about a place or a biographical notice in Onomasticon Arabicum) but are not actually URIs identifying the same kind of entity (a bibliographic resource URI vs. a place or person URI). See the description field below.

Exports into other formats (such as RDF) might be provided in this repository, but for now, the canonical data is in the TSV spreadsheets and other formats are derivative.

Example

uri* related-uri* according-to* description
https://usaybia.net/person/32 http://viaf.org/viaf/81975562 https://usaybia.net is probably the same as
https://usaybia.net/person/32 https://dx.doi.org/10.1163/1573-3912_ei3_COM_24150 https://usaybia.net is the primary entity described by

* indicates required fields.

uri*

required: The URI of the entity (person, place, etc.) in your system that is being related to another URI.

related-uri*

required: The URI of the related entity in another system.

according-to*

required: A URI identifying the project, team, or person responsible for asserting the relationship between the URIs in the first two columns.

description

optional: An informal description of the relationship between the two URIs, for example, "is probably the same as" or "is the primary entity described by". This provides guidance to other projects who may be trying to align their URIs with the ones mentioned. It may also be the basis for creating RDF triples in the future. Please try to keep your descriptions consistent for rows describing the same relationship.

How to contribute aligned URIs

  1. Create a new branch (if you're a member) or fork this repo. To become a member, please contact one of the organization owners.
  2. Add a new TSV file to the relevant folder, named [yourprojectnameorwebsite].tsv, with the columns mentioned above.
  3. Create a pull request to merge your updated spreadsheet into the master branch.

Guidelines

  1. If a URI has multiple related URIs, add a separate row for each relationship.
  2. Please don't include URIs that don't have any related URIs (blank cells). They aren't useful here unless they're matched.
  3. At least one URI in each row (yours or the related one) should resolve to a URL with sufficient information to identify the entity.
  4. Don't add names or other non-URI identifiers to the spreadsheet. This is to make sure all identifications are done on the basis of the URIs themselves, not extra information in the spreadsheet.

How to use and contribute non-aligned URIs

If you have a list of entities for which you don't yet have links to related URIs, you can include these as a TSV in the non-aligned-uris folders for each entity type. We're still working on what should be included in these (at minimum your URI and some kind of label), but you can see an example at https://github.com/Hist-ME/URIs/tree/master/places/non-aligned-uris.

Questions?

Please file a new issue. If you believe a relationship in the spreadsheet is incorrect, please file an issue mentioning the relevant URIs and the line number in the spreadsheet.

Discussion

Discussion for members of the Hist-ME Github organization is conducted on the team discussion page. If you're not a member and would like to be, please file an issue.