Wikidata:Property proposal/review of

From Wikidata
Jump to navigation Jump to search

review of

[edit]

Originally proposed at Wikidata:Property proposal/Creative work

   Done: review of (P6977) (Talk and documentation)

Motivation

[edit]

Women Writers in Review is a collection of 18th- and 19th-century reviews, publication notices, literary histories, and other texts responding to works by early English-language women writers. It is published by the Northeastern University Women Writers Project with support from the Digital Scholarship Group at Northeastern University Libraries, and grant funding from the National Endowment for the Humanities. A digitized evaluation of the reception of these works will enrich the understanding of the data, not just for this collections, but other scholarly collections from around the world as well.. Rosiestep (talk) 21:56, 2 May 2019 (UTC)[reply]

Discussion

[edit]
An example would be: review score (P444) of "Review of A View of Religions" (Q????????) which is a review of "A View of Religions" (Q38227783). What this is meant to do is capture the relationship of the review score (e.g. a rating) of --> the periodical review of --> a literary work. The "review score" or any other associated property we listed would belong with the review and not the work it reviews.
Some facts and definitions regarding Women Writers in Review project:
There are 74 authors, who have published 198 works, which were reviewed by 112 different periodicals.
These reviews include: literary reviews, theatrical reviews, and so forth.
Review scores (ratings) include: Mixed, Very negative, Somewhat negative, Very positive, Somewhat positive, and Neutral or no evaluation.
Reviews are defined as: TEI-encoded transcriptions of individual texts, including data about their sources, and pointers to any referenced works or authors
Sources are defined as: or the historical periodicals/books/etc. in which the reviews were originally published
Works (shorthand for “referenced works”), or the distinct textual creations which are referenced in the reviews

@Rosiestep:, if these work for you, can you update the examples in your proposal? - PKM (talk) 23:00, 8 May 2019 (UTC)[reply]

if User:Rosiestep is happy for this to be what her proposal becomes -- I think it is not what she originally had in mind. @PKM: It would seem that we are clearly identifying this as a proposal for book (film, music, art, etc) reviews, rather than for a review article (Q7318358) or any other kind of review. The specificity may well be a good thing. Are we relaxed that main subject (P921) works for the rest? Jheald (talk) 00:20, 9 May 2019 (UTC)[reply]
See, that's exactly what I meant. If the intention was for book reviews (as hinted by the properties mentioned in the proposal), I'm fine with it, but it's hard to tell because of a lack of examples in the proposals (I have tagged multiple book reviews, I could fetch some!) and what amounts to overdescribed purpose. If the property is actually intended for review article (Q7318358) rather than book review (Q637866) (we don't seem to have a "movie review" item...), then this proposal is exactly what main subject (P921) is meant to handle. Circeus (talk) 00:55, 9 May 2019 (UTC)[reply]
Hello all. Thanks for your feedback and sorry for making things confusing. I am traveling but will be able to circle back to this tonight. Wskent, if you are in Boston by tonight, let's connect, ok? --Rosiestep (talk) 12:40, 9 May 2019 (UTC)[reply]

──────────────────────────────────────────────────────────────────────────────────────────────────── Clarify Planned Use: model the review score (P444) of literary criticism (Q58854) of pre-XX-c women's written work (Q47461344). I'd appreciate any advice on how to proceed or re-write this proposal. --Rosiestep (talk) 15:18, 10 May 2019 (UTC)[reply]

@Rosiestep: Okay, I think what you're talking about above is not a property proposal. A property proposal, in the sense of property proposal pages like this one, is quite a specific thing: namely a proposal for a new "verb" to go into statements of the form <subject> <verb> <object> , to allow one to say something about a particular Wikidata item (the subject).
The discussion you seem to be wanting is much broader, namely how to model "Women Writers in Review" more generally, and the texts contained within it.
I propose first to re-write the proposal above to be just about the verb "is a review of", since that is the kind of specific relation that these proposal pages are for, and seems to have attracted quite a lot of support above. --YesY Done.
Then I propose to create an item for "Women Writers in Review", and some model items for some of the review articles there that you've cited above, that will act as a basis that we can then consider and critique to see whether they could be expressed better. YesY Created: Women Writers in Review (Q63683777)
We might also consider creating a WikiProject page for "Women Writers in Review", as a central place to present summary statistics, and for discussion about data modelling and progress.
I hope that would seem agreeable to you. Best regards, Jheald (talk) 20:36, 10 May 2019 (UTC)[reply]
I have created some items: The Belfast Monthly (1810) on Aikin’s Epistles on Women (Q63684439), The Belfast Monthly Magazine (Q63684444), Epistles on Women (Q63685444), Epistles on Women (Q63685268). They're not perfect, but they're sufficiently developed they can give us a solid enough basis for discussing how best to represent such sorts of items, relationships between them, and different aspects of the Women Writers in Review collection.
If the item for every book review includes the statement collection (P195) = Women Writers in Review (Q63683777), then it will be easy enough to write queries to analyse any collective aspects of the items in the collections; and the works they are reviews of; and those works' authors -- once items are created for each of the constituents of those sets of things. Jheald (talk) 22:29, 10 May 2019 (UTC)[reply]
The part of your project that I am least sure how to model is that corresponding to review score (P444) -- how highly the work was rated by the review.
Conventionally this is a property that sits on the item for the work, see eg Q2377#P444: so one way to represent the WWiR assessment of the review might be:
Epistles on Women (Q63685444)review score (P444)"5/5"
with qualifiers:
review score by (P447) = The Belfast Monthly Magazine (Q63684444)
publication date (P577) = 1810
sourcing circumstances (P1480) = <WWiR assessment>
(better than determination method or standard (P459) ?)
statement is subject of (P805) = The Belfast Monthly (1810) on Aikin’s Epistles on Women (Q63684439)
and referenced by
reference URL (P854) = https://www.wwp.northeastern.edu/review/reviews/aikin.epistleswomen.belfastmonthly.1
retrieved (P813) = 10 May 2019
One thing that I have a bit of unease about is that P444 requires a numerical score -- but the original review didn't give a numerical score. An assessment of the review is made by WWiR, but that's not actually numerical either -- the different values for reception there [1] are in fact "very negative", "somewhat negative", "mixed", "somewhat positive", and "very positive", plus "neutral or no evaluation". If anything, this is more like a scale of -2 to +2, if one should represent it numerically at all. Is it fair and/or appropriate to represent it as a score from 1 to 5 ? I am not sure, even if a P1480 qualifier arguably gives some cover.
But I do like the way that this would group all the review scores together under the reviewed item. In fact under this approach you wouldn't even need to create items like The Belfast Monthly (1810) on Aikin’s Epistles on Women (Q63684439) unless you particularly wanted to, or unless there was particular additional information about the review that you wanted a place to record. And you could use the P1480 tag to gather together reviews from the WWiR project for query purposes.
Any thoughts from people on this modelling? Jheald (talk) 23:22, 10 May 2019 (UTC)[reply]
Neither determination method or standard (P459) nor sourcing circumstances (P1480) have any business being there. What you really wants (as far as I can tell) is reference has role (P6184)book review (Q637866). Circeus (talk) 15:05, 11 May 2019 (UTC)[reply]
@Circeus: It seems that, fundamentally, User:Rosiestep is interested in recording the positivity or negativity of the review, as assessed by the WWiR project. It does seem to me that review score (P444) is our best bet for doing that. But as the original 1810 review didn't come with the helpful statement "Blockbuster -- don't miss: 5* ", it seems to me we do need to qualify any such 5/5 rating with something to point to where that number has come from, what it may mean, and how it has been assessed -- i.e. to link it to an item standing for the WWiR assessment model and methodology. IMO determination method or standard (P459) would be legitimate, because the WWiR methodology is how the 5/5 has been arrived at. But I slightly prefer sourcing circumstances (P1480) as indicating this statement has particular issues associated with it.
This way also has the advantage that one can see how positive or negative all the reviews for a work were, all at a glance in a single place. (And it transfers easily to infoboxes, which is probably why we instituted this approach in the first place.
It's a fair point that it's reductive and uncultured to try to reduce what may be a subtle and engaged discussion of the work to a single numerical score. But it's useful, and that's effectively what WWiR have done, so I don't see any reason not to reflect it here. It's equally true that there will be reviews that we don't have an external assessment of positivity for; and even in the WWiR corpus there are items like notices of works, that give no effective view of them. In such cases we could still use the above data model, even with no numerical score, by setting review score (P444) = somevalue or P444 = novalue and using the same qualifiers. Alternatively, it has been suggested in the discussion at WT:BOOKS to use described by source (P1343) = the magazine that carried the article, with statement is subject of (P805) pointing to an item for the review itself, to which could be added object of statement has role (P3831) = book review (Q637866) along the lines of what you suggest. Jheald (talk) 16:06, 11 May 2019 (UTC)[reply]
My bad. I thought you were discussing actual review scores such as those documented by Template:Video game reviews (Q6354287) (as opposed to stuff like Metacritic (Q150248)'s metascore). Circeus (talk) 23:40, 11 May 2019 (UTC)[reply]

@Wskent, ديفيد عادل وهبة خليل 2, Circeus, Rosiestep, PKM, NMaia: @Jheald, Kolja21, Jura1, Yair rand: ✓ Done: review of (P6977). − Pintoch (talk) 17:35, 3 July 2019 (UTC)[reply]

@Pintoch: Thanks for that. Is it possible to move it up, perhaps to occur between genre (P136) and author (P50) on the page? Jheald (talk) 17:43, 3 July 2019 (UTC)[reply]
@Jheald: I am sure it can be done, you can request it at MediaWiki talk:Wikibase-SortedProperties (I will not do it directly as I am not familiar with handling edit requests for that page.) − Pintoch (talk) 09:21, 6 July 2019 (UTC)[reply]