Wikidata:Property proposal/uses software
uses software edit
Originally proposed at Wikidata:Property proposal/Creative work
Withdrawn
Description | This property will indicate software used in a research study as mentioned in a scientific article |
---|---|
Data type | Item |
Domain | scholarly article (Q13442814) |
Allowed values | software (Q7397) |
Example | |
Planned use | I would like to add this property to scientific articles if the authors state the software used in the study somewhere in the text of the article. |
- Motivation
Creating relationships between the items for scientific articles and items for software will allow us to present more structured data about scientific publications. YULdigitalpreservation (talk) 18:01, 15 June 2017 (UTC)
- Discussion
- Support This property could correspond to a "usedSoftware" property in the DMS ontology that has recently been suggested. https://github.com/mesbahs/DMS/blob/master/dms.owl I am a bit concerned about the name. If the items where it should be used should be confined to scholarly articles and similar (should it?) could "uses software" be misunderstood to a more general property, e.g., an iPhone (Q621427) would "uses software" iOS (Q48493) (similar to P306). Could/should we do anything about this? Call it "used software"? – The preceding unsigned comment was added by Fnielsen (talk • contribs) at 21:39, 15 June 2017 (UTC).
- Oppose use uses (P2283) and cites work (P2860) instead. -- JakobVoss (talk) 07:39, 16 June 2017 (UTC)
- Comment There is one thing that annoys with uses (P2283): The paper itself is not what uses (P2283) the software (Q7397). Rather it is the experiment that the paper describes that uses (P2283) the software (Q7397). — Finn Årup Nielsen (fnielsen) (talk) 13:23, 16 June 2017 (UTC)
- Comment So to be precise the experiment should have it's own item. An alternative would be a very specific property such as "software used in the experiment described in this work". Why can't we view paper=experiment unless there is an item for the experiment? It's not obvious where experiment ends and writing begins, isn't it? For instance Project Jupyter (Q18633895) can be used for both. -- JakobVoss (talk) 07:07, 19 June 2017 (UTC)
- As a practical matter I don't think it's good to require every paper that's described with "usedSoftware" to have a linked experiment. There will be many cases where it's not easy to create items for experiments automatically. ChristianKl (talk) 21:35, 30 October 2017 (UTC)
- Comment So to be precise the experiment should have it's own item. An alternative would be a very specific property such as "software used in the experiment described in this work". Why can't we view paper=experiment unless there is an item for the experiment? It's not obvious where experiment ends and writing begins, isn't it? For instance Project Jupyter (Q18633895) can be used for both. -- JakobVoss (talk) 07:07, 19 June 2017 (UTC)
- Comment There is one thing that annoys with uses (P2283): The paper itself is not what uses (P2283) the software (Q7397). Rather it is the experiment that the paper describes that uses (P2283) the software (Q7397). — Finn Årup Nielsen (fnielsen) (talk) 13:23, 16 June 2017 (UTC)
- @YULdigitalpreservation: Given your examples it seems like allowed value should be software and the domain should be scientific papers. Can we get more clear about what the domain is supposed to be? ChristianKl (talk) 22:40, 17 June 2017 (UTC)
- @ChristianKl: Thanks for suggesting those changes, they make sense to me. I've revised the proposal. YULdigitalpreservation (talk) 13:47, 19 June 2017 (UTC)
Support with the constraint that a quotation (P1683) from the paper supporting the claim be present in each use of this property. Mahir256 (talk) 21:29, 19 June 2017 (UTC)- That suggestion about using the quotation (P1683) seems workable to me. YULdigitalpreservation (talk) 12:12, 13 July 2017 (UTC)
- Oppose, separate item with uses (P2283) is the best suggestion above. Multiple papers can discuss same experiment. d1g (talk) 21:46, 19 August 2017 (UTC)
- This also solves problem(s) when one paper describes 8 experiments and 4 different programs and other similar complexity. d1g (talk) 21:48, 19 August 2017 (UTC)
- Support @JakobVoss: I don't think it makes much sense to tell people to fill uses (P2283) and cites work (P2860) with the same value. This property could subclass both. Having a property with similar semantics as the one in the DMS is valuable for data exchange. ChristianKl (talk) 09:49, 30 October 2017 (UTC)
- A paper can have any combination of uses (P2283) and cites work (P2860) depending on whether it uses and/or cites the software or not. The proposal is about "software used in a research study and mentioned in the paper of the study" not "software used in a research study and cited in the paper of the study", right? -- JakobVoss (talk) 14:12, 30 October 2017 (UTC)
- @YULdigitalpreservation, JakobVoss, ChristianKl: since cites work (P2860) is now one of the most used wikidata properties, I
Supportcreating subproperties that more explicitly identify these relationships - can we call this "cites software" though to avoid some of the problems mentioned above that would apply with the word "uses" in the label? ArthurPSmith (talk) 21:28, 30 October 2017 (UTC)- Maybe "describes software usage" or "used software"(to be more in line with DMS)? It seems like "cites" should definitely be subclassed. Should "uses" given that the item is on the paper? ChristianKl (talk) 21:32, 30 October 2017 (UTC)
- Two questions: Is cites work (P2860) and its subproperties meant to cover only explicit citations or also cases where a paper, software, person etc. is informally mentioned in a text? Do you expect the proposed property to be limited to scientific articles when its label indicates a wider usage (e.g. a company or project uses software)? -- JakobVoss (talk) 07:12, 31 October 2017 (UTC)
- Oppose in favor for the more generic proposal Wikidata:Property proposal/used in the related research project --Pasleim (talk) 11:20, 2 November 2017 (UTC)
- @YULdigitalpreservation, JakobVoss, ArthurPSmith, D1gggg: @Mahir256, Fnielsen: We now have describes a project that uses (P4510), do we still need this one? ChristianKl (✉) 14:32, 12 November 2017 (UTC)
- Agreed, I have struck out my support for this above. ArthurPSmith (talk) 21:00, 12 November 2017 (UTC)
- I am happy to use describes a project that uses (P4510) and we can close this proposal. YULdigitalpreservation (talk) 17:29, 14 November 2017 (UTC)