Wikidata:Property proposal/MxM xref
Jump to navigation
Jump to search
mix'n'match temporary cross-reference[edit]
Originally proposed at Wikidata:Property proposal/Creative work
Withdrawn
Description | statement generated by mix'n'match to sync items lacking "main subject" (P921). Once a catalogue is completed, its statements will be removed. Generally not added manually. See help page for details. |
---|---|
Represents | item lacking "main subject" (P921) about the subject of the statement, e.g. "biography of John Doe" on the item about John Doe |
Data type | External identifier |
Template parameter | n/a |
Domain | varies from one catalogue to another |
Allowed values | Q\d+ |
Example 1 | changes, theoretical sample: Arthur S. Keats (Q89185969) → Q46360334 |
Example 2 | MISSING |
Example 3 | MISSING |
Source | Mix'n'match (Q28054658) |
Planned use | similar to P2536 is used by see Help:Add main subject with Mix-n-Match |
Number of IDs in source | varies |
Formatter URL | https://www.wikidata.org/wiki/$1 |
Robot and gadget jobs | Reinheitsgebot syncs with Mix'n'match (Q28054658) |
See also |
|
Motivation[edit]
Help:Add main subject with Mix-n-Match explains the steps of using Mix'n'Match (MxM) to identify items that are values for main subject (P921).
Initially, I used Sandbox-External identifier (P2536). That works mostly, but could use a few improvements. To avoid problems with other experiments, this proposal for a temporary property. Also, a dedicated property would make it easier to include custom code in MxM.
Note it's temporary in the sense that items for a given catalogue only use it for some time, but it's a permanent property as it can be used by other catalogues later. (Add your motivation for this property here.) --- Jura 20:35, 1 April 2020 (UTC)
Discussion[edit]
- Comment Maybe I'm missing something here, but can't one just use main subject (P921) directly? This seems like an abuse of mix n match... ArthurPSmith (talk) 18:22, 2 April 2020 (UTC)
- How would you do it? --- Jura 18:23, 2 April 2020 (UTC)
- @Jura1: Assuming I understand the idea here, let's say I have some titles of items that are "about" other items. I'd put those titles and their QID's into a csv file, load it into OpenRefine, tweak the titles to pull out likely label matches, then do OpenRefine's reconcile process to match those labels to Wikidata items, then dump out those title QID and matched QID pairs and load into Quickstatements to add the main subject (P921) statements. Or there might be a way to do it directly in OpenRefine. ArthurPSmith (talk) 17:55, 3 April 2020 (UTC)
- I haven't tried OpenRefine's matching recently, but essentially, I would have to do all of them myself ;) Also, I couldn't see what's matching in other MxM catalogues. --- Jura 18:12, 3 April 2020 (UTC)
- @Jura1: Assuming I understand the idea here, let's say I have some titles of items that are "about" other items. I'd put those titles and their QID's into a csv file, load it into OpenRefine, tweak the titles to pull out likely label matches, then do OpenRefine's reconcile process to match those labels to Wikidata items, then dump out those title QID and matched QID pairs and load into Quickstatements to add the main subject (P921) statements. Or there might be a way to do it directly in OpenRefine. ArthurPSmith (talk) 17:55, 3 April 2020 (UTC)
- How would you do it? --- Jura 18:23, 2 April 2020 (UTC)