codecheckers / codecheckers/register

Link the 8 checked works that have no DOI to Wikidata

Open
#212 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Makefile
Stars
6
Forks
3
Avg merge
3m
Merged PRs (30d)
1

Description

Follow-up to #50, where everything else is done.

124 of the 132 certificates state `review of` ([P6977](https://www.wikidata.org/wiki/Property:P6977)) on the work they checked. The remaining 8 cannot: their `paper.reference` is not a DOI, so there is no item to point at, and the export deliberately does not create one - an item with no re-findable identifier would be created again on every run.

| Certificate | Reference | Looks like |
|---|---|---|
| [2020-008](https://codecheck.org.uk/register/certs/2020-008/) | `web.archive.org` snapshot of a CMMID preprint PDF | archival URL, no DOI |
| [2020-009](https://codecheck.org.uk/register/certs/2020-009/) | `cmmid.github.io` PDF | live URL, already at risk of rotting |
| [2022-001](https://codecheck.org.uk/register/certs/2022-001/) | UNESCO IIEP publication page | may have an item or an ISBN |
| [2024-014](https://codecheck.org.uk/register/certs/2024-014/) | Overleaf read link | not citable, needs a real deposit |
| [2024-025](https://codecheck.org.uk/register/certs/2024-025/) | `pure.tue.nl` PDF | repository record may have a DOI |
| [2025-002](https://codecheck.org.uk/register/certs/2025-002/) | ACL Anthology page (LREC 2022) | very likely already on Wikidata |
| [2025-004](https://codecheck.org.uk/register/certs/2025-004/) | TU Delft repository record (uuid) | repository record, may have a DOI |
| [2025-025](https://codecheck.org.uk/register/certs/2025-025/) | raw GitHub HTML report | not citable, needs a real deposit |

What can be done, in order of value:

1. **Find the item that already exists** and add `review of` by hand - 2025-002 and probably 2022-001.
2. **Get a DOI into the register** where the work can still have one (2024-014, 2025-025, possibly 2024-025 and 2025-004). This fixes it at the source; the next export picks it up with no modelling change.
3. **Create the work item from an archival URL** for what will never have a DOI (the two CMMID reports). Technically feasible - `haswbstatement` indexes URL-valued properties, so an item created with its archived URL can be re-found on the next run - but only worth doing for hosts that do not move.

Not an option: pointing `review of` at a URL. It is `wikibase-item`, and so are the alternatives that were considered (main subject [P921](https://www.wikidata.org/wiki/Property:P921), object of statement has role [P3831](https://www.wikidata.org/wiki/Property:P3831)); a URL statement qualified to mean "reviewed" would make a claim about the certificate that is not true, violate the properties' allowed-qualifiers constraints, and still not join certificate to work in any query.

Contributor guide

No contributing guide indexed for this repository

Research direction

Start by auditing the eight certificate references listed in this issue and checking whether corresponding Wikidata items already exist, especially for 2025-002 and 2022-001. Read the follow-up context from #50 and review property P6977 before deciding which records need manual links, source DOI fixes, or archival-URL items. Done means each case has a documented resolution that the export can represent without creating duplicate items.

Written by the indexing model from the issue text.

Assessment

Domain
data
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.