Wikidata talk:WikiCite/Citation Typing Ontology

From Wikidata
Jump to navigation Jump to search

Manually adding CiTO annotations

[edit]
Moved from Wikidata talk:WikiProject Source MetaData/Archive 6#Manually adding CiTO annotations

For the uninitiated, here's some links to useful information about Citation Typing Ontology (Q44955364) annotations:

A recent paper by James L. Powell (Q16104533) (Premature rejection in science: The case of the Younger Dryas Impact Hypothesis (Q110444998)) contains a list of articles that cite An independent evaluation of the Younger Dryas extraterrestrial impact hypothesis (Q24642105) as authority. I went ahead and manually added CiTO annotations to the ones he mentioned, and added a query that generates a similar list to WikiProject Younger Dryas impact hypothesis.

I think it's a great use case for CiTO but it got me wondering: is adding these annotations manually (after publication) an acceptable thing to do? The level of misuse would probably be very low but mistakes could happen, especially for unreferenced statements. I personally think that the benefits far outweigh the potential risks but I wanted to hear from others before going too far and adding any more. Thanks in advance for any thoughts on this! Aluxosm (talk) 19:38, 26 January 2022 (UTC)[reply]

@Aluxosm:, thanks for doing this and asking this question. I have been thinking about this too, and have a personal collection of post-publication CiTO annotations too. I think key is just to allow people to track the authority of the source of annotation. So, I would suggest to upload the annotation to Figshare (Q17013516), Zenodo (Q22661177), or similar and cite that as reference. The metadata in that data repository provide the info people need to decide on the history of the data and how to interpret it. This is the approach I want to take too, but could not find time for yet. --Egon Willighagen (talk) 06:51, 27 January 2022 (UTC)[reply]
@Egon Willighagen: No worries, glad to hear I'm on the right track! I totally agree that some kind of reference is needed when the annotations don't originate from the source. For the ones I mentioned above, I referenced Powell (2022).
For example, on No evidence of nanodiamonds in Younger-Dryas sediments to support an impact event (Q24606726), I added:

cites work
Normal rank An independent evaluation of the Younger Dryas extraterrestrial impact hypothesis edit
has goal cites as authority
1 reference
stated in Premature rejection in science: The case of the Younger Dryas Impact Hypothesis
quotation Instead, as shown in Table 2, right up to the present day many scientists have embraced the results of Surovell et al. to cast doubt on the hypothesis.
add reference


add value


I think I see what you're saying about uploading then citing Zenodo, but I'm not sure how efficient that would be if you only wanted to add a single annotation at a time. Could a simpler alternative (that would allow adding these one-by-one) look something like the following?
With these statements going on Premature rejection in science: The case of the Younger Dryas Impact Hypothesis (Q110444998):

cites work
Normal rank An independent evaluation of the Younger Dryas extraterrestrial impact hypothesis edit
has goal critiques the cited work
1 reference
based on heuristic inferred from prose
quotation Since Firestone et al. showed dispositive photographic evidence that the microspherules exist at the YDB at Blackwater Draw and the other sites, we can only conclude that Surovell et al. failed to sample the YDB and/or erred in their procedures.
(↑ try not to quote more than is necessary to verify the claim)
add reference


add value


Apologies if I've misunderstood what you're saying, thanks for the help! Aluxosm (talk) 12:08, 27 January 2022 (UTC)[reply]
The above solutions is very clear in how the statement is supported. It touches on the discussion whether Wikidata is a primary source or a secondary source. The above looks great to me! The only advantage of putting the data on Zenodo, is that on Zenodo you can further detail how the statements were pulled together. --Egon Willighagen (talk) 06:56, 28 January 2022 (UTC)[reply]
@Egon Willighagen: Nice! Could you ping me if you find the time to do one of yours, I'm interested to see more about how you would implement the Zenodo option. Thanks! Aluxosm (talk) 17:05, 29 January 2022 (UTC)[reply]
This does not appear to be supported by Help:Sources. --- Jura 19:58, 28 January 2022 (UTC)[reply]
@Jura1: Hmmm, do you mean because of this: "In some cases sources are not required: When the item itself is a source for a statement." I've taken that as more of a guideline and in the few cases where I've needed to reference the item itself I've just left out the stated in statement as it's implied. If you meant to talk about the first example statement ↑, I guess it wouldn't be needed if we could reference the statement as I did in the second ↑. Feel like I might have got lost here 😬. Aluxosm (talk) 17:05, 29 January 2022 (UTC)[reply]
  • Unsure about importing quotations I love the project. There are lots of comments that I could make, but for now, I encourage you to go forwarding with modeling examples. One aspect which I question though is importing the quotations. Wikidata uses CC0 copyright licensing and this has to apply globally. There is no provision for fair use or exceptions about copyright. Since Wikidata is global, the copyright needs to be open according to the rules of every country. I am not sure what precedent already exists, but I think importing quotations like this is always problematic and often not permissible. I could be off. Please ask around for precedents elsewhere. As an alternative, you can still say that the determination is from a quotation, but actually copying the quotation in Wikidata may not work. If you do this anyway, and there is a determination that this is not allowed, then there may be mass deletion of the quotations. Please get other opinions as I may be incorrect about best practices. Bluerasberry (talk) 20:58, 15 February 2022 (UTC)[reply]
@Bluerasberry: Are you suggesting that quotes should only be taken from public domain sources? I did check before I started using quotation (P1683) elsewhere (due to the same concerns) but there were/are no restrictions listed on the talk page and the advice at Wikidata:Verifiability seems to suggest that fair use does apply. For reference, the longer quote in the second example above is 41 words (out of the suggested 200 max). I could see this potentially being an issue for single page articles with lots of references (you could end up replicating a significant proportion if not careful), but other than that, this should be okay. It's a fair concern though, keeping the quote to a minimum and only using them when needed is good advice. Aluxosm (talk) 06:42, 16 February 2022 (UTC)[reply]
@Aluxosm: I think everyone should be wary of any quotations. Wikidata is a CC0 project and while that usually is equivalent to public domain, there are circumstances when it is not. Some German Wikidata editors are fond of saying that in Germany there is no public domain, so they call for CC0. No one wants to slow down your project but at the same time, I would not want you to put time into something that may be deleted. If there is a need to import sentences or phrases then probably we should organize a general discussion about the Wikidata policy on quotations for this and other use cases. Or perhaps that discussion already exists somewhere, and there is already an answer. Bluerasberry (talk) 12:18, 16 February 2022 (UTC)[reply]
@Bluerasberry: Not sure how much more I can add to be honest, the current guidelines on quotations align with my thoughts. Requiring quotes to be of the same license as the Wiki they're added to would hobble every single Wikimedia project; this is definitely one for the lawyers! There are currently over 100,000 uses of quotation (P1683) so it might be worth opening a discussion on the property's talk page if you think there could be an issue here. Pinging Thepwnco, who wrote the guidelines, and Matěj Suchánek, who marked them as 'outdated'. Cheers! Aluxosm (talk) 13:21, 16 February 2022 (UTC)[reply]
I think I marked it as outdated because it referred to the deleted property P387 and mentioned some related configuration. --Matěj Suchánek (talk) 13:57, 18 February 2022 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata_talk:WikiCite/Citation_Typing_Ontology&oldid=1949765691"