Wikidata talk:WikiProject Cultural heritage

To dos[edit]

Collect further case reports; make a synthesis of collected reports
Develop a model WikiProject for one of the heritage related areas. Update: See WikiProject Heritage institutions --Beat Estermann (talk) 07:48, 1 February 2017 (UTC)[reply]
Collect show cases for the use of the data in Wikipedia and beyond
Improve generic instructions for data ingestion and data modelling
Create an overview table of different types of heritage data, present coverage, and coverage that should be reasonably aimed for (cf. notability criteria)
Brainstorm the project's success factors
Check the navigation table for completeness
Propagate the navigation table to the various cultural heritage related WikiProjects Done
Come up with further tasks and add them to this list
...

--Beat Estermann (talk) 23:40, 30 November 2016 (UTC)[reply]

It would be good to indicate the tools that help people understand the data and its structure. One tool that is unsurpassed is Reasanator. It provides at a glance understanding on items something that cannot be had in any other way. Thanks, GerardM (talk) 05:48, 1 December 2016 (UTC)[reply]

Redundant[edit]

Did you really need to take over the old name, breaking all links, and to create yet another project with a confusingly similar name? --Nemo 08:32, 1 December 2016 (UTC)[reply]

Hi Nemo, the scope of the former WikiProject "Cultural heritage" did not correspond to its name, as it was exclusively about built heritage. Furthermore, there are three projects on built heritage that will need some coordination in my opinion and some overhauling of the project pages: The one about ingesting monuments data, the one about wikifying WLM, and the Connected Open Heritage Project run by Wikimedia Sweden. For now, I have listed them as "Related Projects" on Wikidata:WikiProject_Built_heritage. I think some housekeeping is needed once in a while if we want to involve more contributors in the tasks related to the ingestion of heritage data. Cheers, --Beat Estermann (talk) 08:42, 1 December 2016 (UTC)[reply]

National Library of Wales[edit]

I just wanted to say that this looks like a great idea, and that i am happy to support this project in any way i can. The National Library of Wales has over 10,000 related Wikidata items and we are constantly looking to add more and expand existing items. We also have the worlds first Wikidata Visiting Scholar. Perhaps some of our work would make a good case study? Like others i am also having difficulty in getting community help with uploads, specifically where a bot is needed. I hope this project might help connect cultural institutions with community members who can help them share with Wikidata. Best Jason.nlw (talk) 08:58, 1 December 2016 (UTC)[reply]

Notability and inclusion of replicable, non-artistic objects in heritage collections - input requested[edit]

I have been approached by a (new to me) group of smaller Flemish museums who want to look at ingesting their collections' data in Wikidata. Most of these museums, however, don't collect visual art objects but rather objects of everyday use, non-unique and replicable objects, etc. Think design objects, bicycles, tiles, textile samples. What is our current position on notability of such items on Wikidata? I know that our notability criteria allow inclusion of any item that is described in serious external sources; and the museums have a shared online collection portal where each item is described. But in a previous project with Flemish museums, we did exclude non-visual art items from one museum (religious heritage items e.g. chalices and chasubles, stained glass fragments) because we didn't consider them 'special' enough. And there's the mythical 'maximum 100 million item cap' that is looming here on Wikidata. Maintenance issues, etc. I'm interested in seeking many perspectives on this issue and hopefully arriving at consensus. WikiProject Cultural heritage has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.

Notified participants of WikiProject Visual arts

Notified participants of WikiProject Fashion Thanks! Spinster 💬 10:00, 29 March 2017 (UTC)[reply]

What makes this more complicated: we are talking about very diverse types of objects.

Visual art items that are produced in series, of which heritage institutions might possess a copy. Think prints and video artworks. As far as I know, we don't have a solid solution for these yet either. Can each copy be a separate Wikidata item?
Notable designs (fashion, furniture...) - not visual art, but objects for personal use - that were produced in smaller or larger series and of which a heritage institution possesses a copy. A Rietveld chair, a pair of 'hoof shoes' designed by Martin Margiela.
Mass-produced products of which a museum possesses one item which may or may not hold special meaning (a typical helmet of an anonymous WWI soldier; a helmet used by a specific well-known soldier)
Mass-produced 'carriers of knowledge' of which an institution possesses one or more copies. Books, photographs...
Unique items for everyday use that may be very special (e.g. a very unique salt shaker) or very mundane (an everyday handmade salt shaker produced by an anonymous artisan)
Objects that can be considered archaeological relicts (pot shards, pieces of flint...), some more special than others

And I'm probably omitting many different categories. Spinster 💬 10:14, 29 March 2017 (UTC)[reply]

Comment I am a big proponent of increasing Wikidata (and Wikipedia) coverage of material culture. Wikidata is in a unique position to provide a central hub for study and understanding of these items across multiple museum collections on many continents. I can imagine all sorts of online presentations of everyday cultural objects in their historical and regional contexts powered by Wikidata and Commons. In my experience, small and special-purpose museums have rich collections of this sort but little money to develop an online presence. As far as dealing with "repeatable" objects, I think the approach used for books ("works" and "editions") offers a solid model for prints, photographs, etc. - PKM (talk) 19:53, 29 March 2017 (UTC)[reply]

I think we'll have to apply subjective criteria, since it's a subjective question.
- Do you mean non-visual art or non-fine-art (eg decorative art)? Non-visual for me means textual (Eg manuscript without illuminations), intangible (traditions), etc.
- I think that interesting decorative arts objects are definitely fair game. Important manuscripts too. Traditions also.
- Pot shards, pieces of flint, scraps of manuscript, etc: should not be in: IMHO they hold only limited professional interest (eg searching for "jug" or Europeana returns tons of shards that can overwhelm you and prevent you from seeing the whole jugs).
- Copies of mass-produced items: if it's an interesting object, we want a copy, but if many second copies are posted, we'd like to prevent this. How to prevent such duplication is a hard unsolved problem, eg libraries have not fully solved it, and they've worked on cooperative cataloging for maybe 20 years. We may also need some extra props along FRBR (eg "copies of this design are are held here and there and there") --Vladimir Alexiev (talk) 06:01, 30 March 2017 (UTC)[reply]

I would approach the issue from the point of view of "completeness": When you start a given inventory on Wikidata - when will it be complete?

First of all, the focus of the inventory should not be on one (or a few) institution(s) alone (i.e. if the answer is: "The inventory is complete when we have described all the items in one institution's collection." - I wouldn't go for it. Have the institution publish their inventory on their own website instead.
Second, the goal of achieving completeness should be reasonable in the sense that you can expect a critical mass of contributors to join the effort in achieving completeness (I think the "Sum of all paintings" project is a good example: the goal is ambitious, but there is also a critical mass of participants ensuring that the project is making substantial progress over time).
Third, the level of granularity in the description of items should be meaningful to (potential) users. I agree with Vladimir Alexiev and PKM in that for mass-produced items, we should aim to describe the items mainly at the level of the product type. FRBRoo V2.4 defines a class "F3 - Manifestation Product Type". It's scope is presently limited to publication products, but I don't really see why it couldn't been applied to other products that can meaningfully be approached from a Work-Expression-Manifestation-Item perspective. If an item is thought to be notable at the level of the individual item, it can of course be described at that level as well.

--Beat Estermann (talk) 06:39, 30 March 2017 (UTC)[reply]

When a serica is produced, it means that the object itself may have an item. It may have all kinds of properties including manufacturer. When multiple GLAM's have one such item, they do not have to have an individual item for the object. An image at Commons can refer to this object. No problem as far as I can see.

Given the way the wikidatification of Commons is likely to go. For every object at Commons there may be an item at Wikidata. So we will have loads of jugs already to start with. So we will get a high influx of data from Commons to begin with. When a GLAM offers us meta data and certainly when we also get pictures, we will have much more data. That may bring us a user interface problem but that is no reason not to have the data. Thanks, GerardM (talk) 08:43, 30 March 2017 (UTC)[reply]

COH Challenge[edit]

Hello! I am just dropping a line to tell you about the COH Challenge that is arranged by Wikimedia Sverige and UNESCO, and that I wrote about at the Project Chat earlier today. I hope some of you want to join! Eric Luth (WMSE) (talk) 13:34, 30 June 2017 (UTC)[reply]

Mapping of South American datasets[edit]

Hi!

I wonder if there is anyone here with good knowledge on cultural heritage in South America and/or Spanish? While working with the Connected Open Heritage project, we have found the following datasets (on Spanish Wikipedia), on South American cultural heritages:

Argentina: Wikidata:WikiProject_WLM/Mapping_tables/ar_(es)
Mexico: Wikidata:WikiProject_WLM/Mapping_tables/mx_(es)
Salvador: Wikidata:WikiProject_WLM/Mapping_tables/sv_(es)
Uruguay: Wikidata:WikiProject_WLM/Mapping_tables/uy_(es)
Panama: Wikidata:WikiProject_WLM/Mapping_tables/pa_(es)

We have, however, some questions on them which are hard to know without background knowledge:

How were the ID numbers of the monuments chosen, i.e. do they come from some official source (in which case, what is it, is it available online?) or were they invented exclusively for WLM
What is the heritage status of the monuments? For example, National Monument of X, Regional Monument of Y. There are two relevant parts
1. Do all the items in a certain dataset have the same heritage status (it's implicit, can be batch-assigned) -- if not, how is it indicated?,
2. Do Wikidata items for the heritage status(es) exist?

Thanks, Eric Luth (WMSE) (talk) 12:29, 12 September 2017 (UTC)[reply]

Hi Eric, try asking your question here: Wikidata_talk:WikiProject_Built_heritage; I guess that's the page most of the WLM people have on their radar. --Beat Estermann (talk) 10:05, 25 September 2017 (UTC)[reply]

@Beat Estermann:: Many thanks! Eric Luth (WMSE) (talk) 12:25, 26 September 2017 (UTC)[reply]

Varying collection sizes for one institution[edit]

A museum states "[our collections] include over 25,000 objects [...] The majority of the 100,000 volumes in our library [...] We hold over 1 million photographs". What's the recommended way to connect the three different numbers to the museum's entry with collection or exhibition size (P1436), given that the library and the photography collection are not separate organisations? Thanks in advance for any advice, MartinPoulter (talk) 10:38, 7 December 2017 (UTC)[reply]

applies to part (P518) ? Jheald (talk) 00:14, 8 December 2017 (UTC)[reply]

Copyright policy of an institution?[edit]

Hi all, on a twitter thread recently, we were discussing how to document institutions that use different open-access/copyright policies for their collections. One proposed application, was using Wikidata (see the Tweet). Have any collections policies been modeled in this way? Do we have the appropriate properties to do this? WikiProject Cultural heritage has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead. Astinson (WMF) (talk) 18:18, 7 March 2018 (UTC)[reply]

WikiProject sum of all paintings has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead. Astinson (WMF) (talk) 18:19, 7 March 2018 (UTC)[reply]

It might be tricky, policies change and are probably not universally applied by most institutions. Data would still be useful. All the best: Rich Farmbrough, 16:02, 8 March 2018 (UTC).

We already face this issue with digital library (Q212805) and biological database (Q4117139) and it seems to be currently represented with copyright license (P275). This could be given qualifiers such as applies to part (P518), start time (P580) or end time (P582)

SELECT ?d ?dLabel ?licenceLabel WHERE {
VALUES ?resource {wd:Q212805 wd:Q4117139}
?d wdt:P31 ?resource.
OPTIONAL{ ?d wdt:P275 ?licence }
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
} ORDER BY DESC(?licence)

Try it!

MartinPoulter (talk) 16:43, 12 March 2018 (UTC)[reply]

I didn't say thanks when this came up previously @MartinPoulter, Rich Farmbrough: We discussed this a bit at Creative Commons confererence, and Wikidata might the tool for keeping track of the landscape for these policies. I have a couple meetings in the next couple months, and update you all, if something more comes from it. Astinson (WMF) (talk) 13:22, 25 April 2018 (UTC)[reply]

Feedback welcome: starting GLAM-related ontology and metadata mapping for Structured Data on Commons[edit]

WikiProject Cultural heritage has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead. Hi all! For those interested in ontologies and metadata schemes around art and culture (GLAM - Galleries, Libraries, Archives, Museums), and around media files and the upcoming structured data on Commons: there is currently a first feedback round (and probe of interest) to start working on metadata and ontology mapping around this topic for Wikimedia Commons. Please check the request for feedback page on Commons, and leave your comments on the talk page there! SandraF (WMF) (talk) 09:59, 25 April 2018 (UTC)[reply]

Help with Lefkara lace[edit]

I am not sure how to record the UNESCO Intangible Cultural Heritage info for Lefkara lace (Q6516672). UNESCO page is here. Can someone help? - PKM (talk) 02:09, 28 June 2018 (UTC)[reply]

I have taken a first stab based on the comments at Talk:Q59544, but it doesn't seem right. Saying a Wikimedia list article is a catalog is bad ontonology; I would think we need separate items for "UNESCO Representative List of the Intangible Cultural Heritage of Humanity" and "UNESCO List of Intangible Cultural Heritage in Need of Urgent Safeguarding" for the catalog item. Thoughts? - PKM (talk) 18:46, 28 June 2018 (UTC)[reply]

What are the best modelled items for your areas of interest?[edit]

Hi all

Over the past few months myself and others have been thinking about the best way to help people model subjects consistently on Wikidata and provide new contributors with a simple way to understand how to model content on different subjects. Our first solution is to provide some best practice examples of items for different subjects which we are calling Model items. E.g the item for William Shakespeare (Q692) is a good example to follow for creating items about playwright (Q214917). These model items are linked to from the item for the subject to make them easier to find and we have tried to make simple to understand instructions.

We would like subject matter experts to contribute their best examples of well modelled items. We are asking all the Wikiprojects to share with us the kinds of subjects you most commonly add information about and the best examples you have of this kind of item. We would like to have at least 5 model items for each subject to show the diversity of the subject e.g just having William Shakespeare (Q692) as a model item for playwright (Q214917), while helpful may not provide a good example for people trying to model modern poets from Asia.

You can add model items yourself by using the instructions at Wikidata:Model items. It may be helpful to have a discussion here to collate information first.

Thanks

John Cummings (talk) 15:42, 17 December 2018 (UTC)[reply]

Adding translations for the project pages[edit]

Hi all! I would like to add the translation markup (and immediately translations into Russian) for this page, as well as for Wikidata:WikiProject Heritage institutions and Wikidata:WikiProject Built heritage (at start), and then for another sub-projects. The ability to translate navigation tabs will also be added.

If there is no objection, then I will start this week. I have experience in performing such work in Wikidata, as well as "translation administrator" rights, so that page corruption is not expected. However, I apologize in advance for any inconvenience.

PS: To @Beat Estermann: - I am in contact with an Olga Barysheva from Russian National Library, and this task was initiated from her request. --Kaganer (talk) 11:09, 27 August 2019 (UTC)[reply]

Describing location of buildings in archaeological sites[edit]

WikiProject Cultural heritage has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.

I am wondering about the best way to model archaeological sites. Take, for example, the Dura-Europos amphitheatre (Q29001887), which was excavated in the archaeological site of Syrian Dura-Europos (Q464266). I notice some buildings linked to the site of Dura-Europos are described as part of (P361) of Dura-Europos while other buildings are described as located in the administrative territorial entity (P131) of Dura-Europos. I see cons to both: the former ("part of") is used more for objects rather than buildings in locations but the latter can be confusing (the building was located in the administrative territory of Dura-Europos in the Roman period" but it could also be described as located in the administrative territory of Al-Salihiyah, Deir ez-Zor Governorate or the Abu Kamal District. So we end up with multiple values). I am wondering what is best practices for talking about blocks or sites within an archaeological site.Valeriummaximum (talk) 20:51, 29 August 2020 (UTC)[reply]

how to model inception for date ranges[edit]

WikiProject Cultural heritage has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.

Notified participants of WikiProject Archival Description

Hello, I'm not sure where to address this question. I am wondering how one would parse a catalogue description like "made 123BCE-256CE". Since inception (P571) requires a point in time rather than a date range, it's not possible to enter this date range. Naturally, earliest date (P1319) and latest date (P1326) would be the best way to model the earliest and latest dates of the date range (and one can) but Wikidata prefers that these properties be used as constraints rather than as main statements. Does anyone have any suggestions on how to work with date ranges to represent the broad range of an object? Valeriummaximum (talk) 12:59, 15 October 2020 (UTC)[reply]

Data modelling conventions of instance of (P31), genre (P136), levels of specificity, and consistency across cultural heritage description on Wikidata[edit]

WikiProject Cultural heritage has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.

On my talk page, I have been addressed by Hjart, specifically about the usage of bust (Q17489160) as either instance of (P31) or genre (P136) for sculptures. I think this issue is pretty interesting, and would benefit from arguments from a larger group of people: it touches upon broader issues for cultural heritage on Wikidata, namely what kind of information we want to place in which parts of the ontology and why, and what level of consistency we want to achieve in descriptions of cultural heritage works across 'art forms'.

Since quite early, the WikiProject Sum of all paintings has in practice used the generic painting (Q3305213) for the vast majority instance of (P31) statements, and has moved (also general) subject matter to genre (P136), for instance religious art (Q2864737) or self-portrait (Q192110).

I tend to agree with that approach for various reasons. We then use instance of (P31) in a high level way for the 'general identity' of an object (i.e. not the object's subject matter), and use other properties like genre (P136) and movement (P135) for more precise, content and subject-based refinements. I think it's a good approach for various reasons:

Ease of querying;
Opening the potential to do faceted search and browsing in a 'cleaner' way (less traversing of subclass trees);
Ontologically 'cleaner' - instance of (P31) is then used for the 'pure essence' of what a creative work IS.

This is why I have been moving certain content-based categories of sculptures - for instance equestrian statue (Q659396), but also the general category of public art (Q557141) (as documented in its WikiProject), under genre (P136). I'm not the only person who does this - I notice that other quite active Wikidatans use this approach. Personally, I think it's good to have a certain level of consistency between various art disciplines as well.

However, I'm curious about other viewpoints. I'm posting this question here because I'm wondering:

Do folks, also those who work on cultural heritage outside painting or sculpture, generally agree with the reasoning above? Do you have any additions, refinements, or improvements? If you don't agree, what are your reasons for disagreeing? Do you use a very different system in 'your' heritage discipline, and if so, why? Do you know of other data models that suggest otherwise, and if so, which are these?
What are your thoughts about consistency across cultural heritage disciplines? Is it worthwhile to actively pursue this?

I think it would be awesome to reach some well-argumented consensus on this, which we can then document and implement better. Very curious to hear what folks think. --Cheers, Spinster 💬 12:27, 15 October 2020 (UTC)[reply]

Is this someone you think should be applied to general categories like equestrian statue (Q659396) or to specific objects as a rule? I notice that a number of equestrian statues are classed both by [instance] and by [genre]. Valeriummaximum (talk) 12:50, 15 October 2020 (UTC)[reply]

I totally agree. The only reason I'm not doing that is an habit (and habits die hard). Cheers, VIGNERON (talk) 13:10, 15 October 2020 (UTC)[reply]

Just to add to this, I am preparing a large batchupload of various kinds of archaeological objects (some of which are busts, statues, frescoes, but also commonware items like dishes, vessels, chairs), see [here for project description] I had intended to just add a claim "instance of" for all of these labels, but following your standard, I think I would need to separate many of these labels as "instance of" claims and "genre" claims, which I would be happy to do if this was the community standard. My main concern is that this distinction might not be intuitive for someone querying. If someone wanted to find all artifacts classed as "vessel" or "altar" or "bust" or "fresco" or "chair", I think they would intuitively search by instance of (P31) rather than genre (P136). It is true that genre (P136) can be applied to anything that is a subclass of work (Q386724) but I do wonder if "genre" stretches the natural meaning when we start applying it to things like "equestrian statue" or "bust" (as opposed to something like "Roman art" or "Monumental Art" which are easier to see as genres), especially if the home gallery catalogues do not use the language of "genre". I absolutely agree there needs to be a clearer standard because clearly people are using "instance" and "genre"--would it be possible to see how this could ease querying?Valeriummaximum (talk) 13:46, 15 October 2020 (UTC)[reply]

Not an expert at all (unless you agree that video games are cultural heritage ;-), but I agree with the general idea. One thought, form of creative work (P7937) could be a good tool as well − as far as I know, it was conceived primarily for written works and extended to musical albums, but it could apply to other domains? Jean-Fred (talk) 08:05, 16 October 2020 (UTC)[reply]

I am also no expert, but I know enough just from poking around various datasets on Wikidata that we should probably set down some guidelines, if only to preserve our own sanity. I see your corrective edit to my recent change on Singraven: bust of Pan (or Satyr?) on a pedestal (Q17596039) and notice you did not revert me, so I assume you are OK with naming things in green spaces as art, though the item was previously instance of (P31) green space (Q22652). Another edit on Singraven: garden table (Q17596052) that I made on the same Singraven (Q18772557) batch was to change green space to garden table (Q65660742) though after looking at the photo I am inclined to call it art as well. So individual objects in green spaces that are sculptures should probably be called sculptures, despite what the heritage designation is? Or should the green space be added back to the P31? I honestly don't know. Jane023 (talk) 09:23, 16 October 2020 (UTC)[reply]

Thanks everyone for your insightful comments (and of course further comments are very welcome). Feel free to add and improve, but I hear some other (quite good) arguments in favor of, and against, certain instance of (P31)/genre (P136) (and other property) distinctions:

Choose conventions that make it as intuitive as possible for people who search and query, also people who are not familiar with the data. I agree that indeed in some cases more specific instance of (P31)'s may be better in this case. Although, as one example, personally I would never look for 'busts' but rather for 'sculptures' as a newbie, as 'buste' is (in my native Flemish Dutch) an old-fashioned and rather specialized term and I expect that in that cultural context more people will look for sculptures in general. So I think cultural and linguistic differences in naming things also play a role here.
Follow closely what, for instance, a cultural institution is saying itself. So, if an institution calls something an equestrian statue in its dataset, then also use instance of (P31) for that here on Wikidata. But then: what about consistency on our side - we risk to end up with very diverse approaches here too? What about decisions we want to make because we want to improve the data (as in Jane's example of the table)? Also, for quite a few cultural objects on Wikidata there may not really be a 'source dataset'; for instance, I've been creating quite a few Wikidata items for artworks in public space that have categories on Wikimedia Commons, but that are not described in very formal databases.

I posted the original question here because I'm really interested in comparing approaches across disciplines (and yes, Jean-Fred, I totally think video games are very important new cultural heritage!) and hopefully, as I said, perhaps getting to more clear guidelines to help all of us. Would there be a way in which we can conveniently and constructively talk about this further without going crazy (because it's so much data to look at)? My first thought would be to create some dedicated wiki pages for this and perhaps also to use something like a shared (Google) spreadsheet to play around with things but maybe others have better ideas? Cheers, Spinster 💬 14:08, 16 October 2020 (UTC)[reply]

To expand a bit on how we do things over at Wikiproject video game: in general we use -instance of (P31)video game (Q7889), and not any subclasses. For "theoretical" reasons (as in, good modeling) and for practical reasons − these subclasses are too porous for that purpose.

For example, we do not use P31 with:

genres, eg action game (Q270948): it’s covered by genre (P136) − and in practice, many games span/blend several genres
the triplet console game (Q6561427)/PC game (Q4485157)/mobile game (Q1121542): it’s covered implicitly by platform (P400) − and in practice, PC games are re-released on consoles, console games get ported to mobile, etc.

I’m also wary of just using what some other database say − sometimes, it’s more of a limitation of their modeling than deliberate. For example, MobyGames (one of the oldest and biggest databases) uses video game compilation (Q16070115) as a genre − but in my view that’s just because they shoehorned that distinct concept into a data model which was meant for games. And we on Wikidata consider compilations as a first-class entity. So maybe the institution vocabulary is very limited (and they shoehorn things around), or their grammar is very limited (for example, they have no field for the artistic movement, so they record it in the main field), and I see limited reason for us to reflect these limitations.

This also seems to be the route taken by Wikiproject Books: using P31=written work, and use form of creative work (P7937) to express novel, play, etc.

Jean-Fred (talk) 16:33, 16 October 2020 (UTC)[reply]

I support setting up page to spell out guidelines. I personally feel that "genre" should be used more restrictively because I am not sure that people would naturally think of "busts" and "equestrian statues" as genres of sculpture but rather as types or instances of sculpture (in a way they are both and so I agree with Spinster, but to someone querying, which is intuitive?). I also think it is important that any guidelines should reflect existing heritage guidelines like CIDOC Conceptual Reference Model (Q624005) or Art & Architecture Thesaurus (Q611299). It concerns me a little that, for example, a page like genre of sculpture (Q18783400) has no references--we can't be using categories and classes to define our ontologies which have no references and might appear opaque to editors. But to reiterate, I think this is an important topic and am thankful to Spinster presenting it so well.

Valeriummaximum (talk) 21:11, 16 October 2020 (UTC)[reply]

I also totally agree that looking at existing standards has to be part of the discussion, and that we need more 'discipline' and accuracy, better sources and referencing for our ontology-level concepts! But then, I just looked at the Europeana Data Model Mapping Guidelines and that one just states 'use terms from a controlled vocabulary' for dc:type of cultural heritage objects and even seems to conflate it with 'genre' a bit. Haha!

As only a smaller group of folks involved in cultural heritage on Wikidata are a member of this specific WikiProject and more voices and participants would be welcome, I will take the freedom to ping a few other active WikiProjects where I think there will be people with interest and good ideas. I apologize for the broad ping but hope that this topic may be useful, and we can form a small group of folks who want to work on this across disciplines. Cheers, Spinster 💬 07:50, 17 October 2020 (UTC)[reply]

Notified participants of WikiProject Visual arts

Notified participants of WikiProject Performing arts

Notified participants of WikiProject Built heritage

I mostly work in the areas of costume and decorative arts. My personal preference is to match P31 to the usage in the contributing collection, which tends to be fairly specific. ~~Here's a list of <instances of> "cup" and its subclasses>.~~

The standard vocabularies in this area are Art & Architecture Thesaurus (Q611299), which is (IMHO) way too granular for Wikidata in some domains, and Nomenclature 4.0 for Museum Cataloging (Q82310767), which is pretty solid. Nomenclature is not yet strongly mapped to Wikidata, but there's a Mix'n'Match catalog and we're working on it.

One challenge to standardizing on a basic P31 and a supporting genre is that there really aren't "genres" of clothing or dishware. These are discussed as "types" and type pretty much means subclass to me. It would be awkward to talk of "walking dress" as a genre of "dress", or "romer" as a genre of "glass". - PKM (talk) 21:31, 17 October 2020 (UTC)[reply]

I wonder if we need different standards for the traditional “fine arts” and for “material culture” objects? PKM (talk) 19:34, 18 October 2020 (UTC)[reply]

I think this discussion would benefit from more comments by GLAM practitioners, and so I posted a note on the Wikidata + GLAM Facebook group. - PKM (talk) 19:48, 18 October 2020 (UTC)[reply]

Sorry, my query above is broken. We have a *lot* more <instances of> subclasses of clothing, only one or two of them added by me, and many of them are specific (here are "walking dresses" https://w.wiki/i7Y). - PKM (talk) 22:51, 22 October 2020 (UTC)[reply]

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘A similar discussion as the one above is now also taking place at the talk page of WikiProject Visual arts, but specifically focused on prints. Spinster 💬 20:57, 23 November 2020 (UTC)[reply]

"repository"[edit]

I was looking for a good item for the sense of the word "repository" when referring to the institution that collects/preserves an object, but I am not finding a good item for this basic GLAM term. Specifically, I am looking for something like definition #2 on [1], which is not about the place where items are stored, but the institution that cares for such items. repository (Q2145117) is about the storage site, not the organization and institutional repository (Q1065413) is too specific for this sense. Similarly, while we have items for organizations like archive (Q166118), any type of organization can be a repository. Other items are even more specific (archival repository (Q66656823)). Anyone have a good item for this, or want to take a stab making one? Dominic (talk) 18:20, 13 July 2021 (UTC)[reply]

Name of Churches in Italian[edit]

FYI discussion in Italian about the use of capitol letters in some names. Some non it-N users have been correcting here and there the word "chiesa"/"Chiesa", so we are trying to dig into the sources and clear up the ambiguity.--Alexmar983 (talk) 17:06, 31 July 2021 (UTC)[reply]

Adding item Type Constraints for Historic Marker Properties[edit]

Cleaning up the current application of historical and commemorative markers... The current practice appears to be are a mix placing the ID on "subjects" of the marker as well as on instances of "commemorative plaque (Q721747)" or subclasses of such". I'm suggesting following the instances of an object about a subject and providing constraints to move practice to this. Given these IDs represent physical objects in the real world and are different in concept than the subject of their marker, the propensities should be references to a subject rather than on the subject themselves. This approach affords cleaner constraint and linking and disambiguates the concepts. This approach also affords distinguishing the location of the sign from the location of the subject when mapped. This means links to an object resolved with the Q in OpenStreetMap (OSM) which is of particular interest due to the Wikimedia maps instance built on OSM as well as other geographic uses would be cleaner to type matching of objects. Queries in Wikidata would be more specific to the object referenced by the Q rather than a conflated Q. Additionally for coordinates known for the marker that marks the location of an event with the relation to the event Q it would be a lead for creating a better coordinate for the event if the location is missing or unknown.

Properties identified so far to be affected (please add any that may have been missed):

Nevada SHPO marker ID (P9515) - Type Constraint added 27 Feb 2022 (no prior type constraint existed) - Constraint Report for P9515
Historical Marker Database ID (P7883) - Type Constraint added 26 Feb 2022 (no prior type constraint existed) - Constraint Report for P7883
Florida Historical Marker List ID (P6567) - Type Constraint added 27 Feb 2022 (no prior type constraint existed) - Constraint Report for P6567
Indiana State Historical Marker Program ID (P9546) - Type Constraint added 27 Feb 2022 (no prior type constraint existed) - Constraint Report for P9546
NC Highway Historical Marker Program ID (P9492) - Type Constraint added 27 Feb 2022 (no prior type constraint existed) - Constraint Report for P9492
Open Plaques plaque ID (P1893) - Type Constraint added 26 Feb 2022 (no prior type constraint existed) - Does not apply to the subject ID - Open Plaques subject ID (P1430), but other IDs might want to replicate having a subject ID approach if an ID is accessible for data integrity workflows. - Constraint Report for P1893
Chi era Costui plaque ID (P9996) - Type Constraint applied since 20 December 2021 - Constraint Report for P9996

TODO:

Cleanup examples of property IDs currently pointing to subjects
Cleanup constraint violations (see links associated to the reports above)
Clarify differentiation relationship between commemorative plaque (Q721747) and information sign (Q6031215)

Questions:

What property should be the required reference to the subject of the plaque? Should the reference to the subject be main subject (P921)?
What additional constraints should an instance of or subclass of "commemorative plaque (Q721747)" require so these constraints can be added to create a model for markers / commemorative plaque
Are there cases where commemorative plaque (Q721747) and information sign (Q6031215) both are true?

Wolfgang8741 (talk) 19:44, 27 February 2022 (UTC)[reply]

With regards to Historical Marker Database ID (P7883) how would a one to "group" rather than individual marker be represented. Would an instance of commemorative plaque (Q721747) with a qualifier of group be appropriate? The grouping was mentioned in the initial proposal. Wolfgang8741 (talk) 00:09, 28 February 2022 (UTC)[reply]

Pinging the property creators of the IDs listed above: @UWashPrincipalCataloger:@Pintoch:@Nomen ad hoc:@Pigsonthewing:@Pamputt:

Seems a bad idea to me. All the examples when the property were created were actual places, and I can't see why we'd want WD to have entries for individual marker plaques in addition to them. Plaques are not statues. There are only 6700 plaques in the US, and 192k historical markers in the HMDB database, it would seem much more useful to add them to the sites they describe than label bits of metal. Vicarage (talk) 10:20, 4 March 2024 (UTC)[reply]

Poperty creation[edit]

Hello , exhibited creator (P10661) —Eihel (talk) 07:20, 19 April 2022 (UTC)[reply]

Use of P527 vs P912 for parts of monuments[edit]

Ongoing discussion (unfortunately the project has surpassed 50 members, so {{Ping project}} cannot be used anymore). --Epìdosis 15:19, 24 April 2022 (UTC)[reply]

@Epìdosis: Current guidance is that has part(s) of the class (P2670) is the property to use for this. But we might want to think whether that is actually what we like. Jheald (talk) 17:07, 24 April 2022 (UTC)[reply]

Proposal of update of format violations (P3503)[edit]

Hi, if interested on P3503's violation, please join the discussion here Property talk:P3503 Elemar (WMIT) (talk) 12:50, 17 June 2022 (UTC)[reply]

New project on Intangible Cultural Heritage[edit]

I have created and worked on a project on Unesco Intangible Cultural Heritage. I have already imported all the items on the UNESCO Intangible Cultural Heritage Lists (Q4435332) but a lot of work remains to be done disentangling items, arranging list articles, translating, cleaning up hierarchies, linking to national lists and possibly even modifying property definitions etc. Please feel free to join / comment in Wikidata:WikiProject Intangible Cultural Heritage and the discussion thread in the Project chat at Wikidata:Project_chat#Looking_for_feedback_on_items_about_intangible_heritage. Cheers, Susanna Ånäs (Susannaanas) (talk) 11:24, 12 July 2022 (UTC)[reply]

OPEN CALL FOR A WIKIPEDIAN IN RESIDENCE AT PAMM[edit]

The Pérez Art Museum Miami (PAMM) is seeking a creative and driven Wikipedian In Residence (WIR) to improve coverage of their collections on Wikipedia, Wikimedia Commons, and Wikidata. PAMM is a modern and contemporary art museum dedicated to collecting and exhibiting international art of the 20th and 21st centuries.

MUSEUM: https://www.pamm.org

The WIR would be helping PAMM establish Wikipedia programming institution-wide. The Wikipedian In Residence will work closely with the digital team at PAMM with access to curators and staff for support and training.

DIGITIAL LEAD: https://www.jaymollica.com

On Wikidata, the WIR will add structured data and expand items about painting objects, artists, collections, installations, exhibits, and events. On Wikimedia Commons, photos of living artists will be taken and added with appropriate categorization. On Wikipedia, the WIR will improve articles about PAMM's artists, their key collections, and the notable art movements of which they are a part. The WIR will also engage community through edit-a-thons, virtual events, and staff trainings.

They WIR will have a dedicated, expert strategic advisor helping them onboard into the role and plan their activities with weekly check-ins and ongoing support.

ADVISOR: https://www.wikiblueprint.com

A good candidate will have some (but not necessarily all) of the desired skills and experience. We encourage a diversity of candidates to apply even if your resume doesn't meet every qualification.

PLEASE APPLY: https://secure5.saashr.com/ta/5CET.careers?ShowJob=302365959

Ocaasi (talk) 22:41, 29 September 2022 (UTC)[reply]

Just a reminder that applications are still open, REMOTE applicants are OK, and a resume/cover letter is all you need to submit.

Apply: https://secure5.saashr.com/ta/5CET.careers?CareersSearch

Ocaasi (talk) 23:02, 18 October 2022 (UTC)[reply]

Proposal for Cultural Good of Croatia ID[edit]

Requesting feedback on the following Wikidata:Property proposal/Cultural Good of Croatia ID.

A little background, the Register of Cultural Goods of Croatia (Q7309359) hosts historical photos and metadata for important buildings and monuments in Croatia. Linking Wikidata to this resource will help import three lists maintained by the Ministry of Culture and Media of the Republic of Croatia (Q3443835) in Croatia:

List of protected cultural goods (Lista zaštićenih kulturnih dobara)
List of cultural goods of national significance (Lista kulturnih dobara nacionalnog značenja)
List of preventively protected goods (Lista preventivno zaštićenih dobara)

There are a few of these cultural sites already in Wikidata. Creating this external identifier will help expand the coverage of these cultural areas. This is my first property proposal. If I'm missing anything in the proposal or not following the expected process for approval, please let me know. - DutchTreat (talk) 11:47, 20 November 2022 (UTC)[reply]

Question about instance of (P31)[edit]

Hello,

I have a question about instance of (P31): Should heritage site (Q358) always be entered for instance of (P31) as well, or should only heritage designation (P1435) be filled in accordingly?

So, for example, at instance of (P31): house (Q3947) AND heritage site (Q358) be entered? Or is heritage site (Q358) then not necessary?

--N8eule78 (talk) 08:24, 20 February 2023 (UTC)[reply]

I don't really like to use monument (Q4989906) or heritage site (Q358) as instances (we have lots of wrong monument (Q4989906) in elements in Spain). IMO, the heritage designation (P1435) is more clear to use because it is an administrative fact and can (should) be referenced with some doc published by the government.

Then we can find a gray zone for elements someone believes or qualifies as a monument but they are not designated by the government. I'm not saying they are wrong, but I really don't know how to model it.

But I'm sure if we are talking about things made with atoms (like house (Q3947)) they can't be instanced also as some human convention (Q367293), a social construct. —Ismael Olea (talk) 09:54, 20 February 2023 (UTC)[reply]

Hello again,

I am specifically concerned with heritage monument in Saxony (Q19413851).

There it is clear that it is a cultural monument (cultural property (Q2065736)). Now the question is whether in these cases "cultural property (Q2065736)" (or "heritage site (Q358)") should additionally be entered at "instance of (P31)"...

A concrete example where this has already been done is pigeon house Falkensteiner Straße 19 (Q49220758). I would like to know if this is wrong or if this is allowed (or should be done).

--N8eule78 (talk) 09:23, 21 February 2023 (UTC)[reply]

To me in that case is perfectly fine to only use the heritage designation (P1435) with value heritage monument in Saxony (Q19413851) to state the monument consideration. No need to use P31 but just describing the physical building. Doing it we would be mixing a physical description (a dovecot) with a convention (Q367293) (a monument). Also it would be a redundant statement because the monumental qualification is unambiguously set with heritage monument in Saxony (Q19413851).

I have a little experience reusing data like these and this is a very straight forward way to identify and process elements. But It's only my point of view.

If I was wrong I'd really love a set of properties restrictions (or ShEx models) available to check Wikidata conventions and quality assurance. —Ismael Olea (talk) 10:47, 21 February 2023 (UTC)[reply]

OK, Thanks! - I changed it on the two objects where I noticed it, and will do so in the future.

--N8eule78 (talk) 11:47, 21 February 2023 (UTC)[reply]

Describing overlapping location data for find-spots[edit]

Hello everyone - I'm working on Wikidata:WikiProject IDEA and we're trying to represent data about the find spots for artifacts, using the information provided in the excavation reports. The objects in question were excavated during the early 20th century and so the find spot information is not as precise as might be expected from modern excavations. Frequently, the reports will say something like 'found on X street, back of Block Y', therefore providing us with an approximate area for the discovery that can be pinpointed to the overlap between the street and excavation block referred to. We've been trying to decide what the best way to represent this information would be. There have been a few suggestions:

1. Use multiple location of discovery (P189) statements, with the addition of the qualifier sourcing circumstances (P1480) and values such as near (Q21818619) for instances where the reports say things like 'near Tower Y'. A potential issue with this option is that we worried it might suggest uncertainty about the find spot, rather than overlapping information (and might not work well with the map visualisation tool currently available with the querying service)

2.Use only one location of discovery (P189) statement (picking the main or most specific information available), and then represent overlapping location data with a qualifier. The selection of a qualifier has been tricky, however. Ideas of what to use include:

i) part of (P361) - concerns with using this as a qualifier however included the symmetry constraint, which would influence the way the locations themselves are represented, and also didn't seem to quite express the right idea (it's not that these areas are a part of one another so much as they overlap each other geographically?)

ii) significant place (P7153) using object has role (P3831) as an additional qualifier with a value like contained within (Q96783935) - main issues with this is that significant place (P7153) doesn't seem to be often used as a qualifier, and that when there is more than one overlapping location described, or information like 'near Tower X' is given, the object has role (P3831) secondary qualifier values might become confused regarding which significant place they pertain to.

iii) including (P1012) - mostly used to qualify quantities, however. We're not sure if this expresses the right idea.

iv) Propose a new property expressing the concept 'overlapping'.

Do people have thoughts or advice about this? Any other possible properties we might have missed or thoughts regarding proposing a new property? Thanks in advance for your help! JASHough (talk) 12:59, 2 June 2023 (UTC) @Ahc84:[reply]

Notified participants of WikiProject Ancient Greece

Notified participants of WikiProject Ancient Rome

Notified participants of WikiProject Archaeology

Excellent issue to raise! Good modeling of find-spots is very important if Wikidata is to become, as I hope, a tool for reconnecting moveable monuments to their original setting. Bravo! My preference would be to choose in each case the option that generates the most accurate imputed coordinates for mapping purposes.

What is the convention for modeling find-spot coordinates? My thought was to use coordinate location (P625) with object has role (P3831) with value location of discovery (Q1291195) but I don't find a significant number of instances of that. JBradyK (ToposText) (talk) 15:19, 2 June 2023 (UTC)[reply]

Thanks for raising this issue @JASHough, and for lending your brain to it, @JBradyK. IDEA is a legacy data project, so in many cases we are dealing with fuzzy findspot locations. For that reason, we have so far shied away from specifying coordinates for findspots since in most cases the site documentation from the 1920s/30s only states that the object was found "in or near xyz building/feature", without any means of assigning an absolute coordinate. We didn't want to imply more specificity than we can substantiate. Our strategy to best reflect the records that we have, therefore, has been to systematically create entities representing all of the substantial archaeological features from the site that are referenced in the literature, and associate all of those entities with geoshapes. Each artifact with a fuzzily recorded findspot will then receive a "location of discovery" (P189) statement that points to the building/features noted in the legacy documentation.

So for instance, "XPapyrus (Q#TBD)-->Location of discovery (P189)-->wall street, Dura-Europos (Q116621880)", but then determining how to qualify that statement to express that the object was found in wall street, but somewhere in the section of wall street that is coincident with a specific city block.

To add to JASHough's list of possibilities, we were wondering too if "Adjacent building" (P3032) might work to qualify such a statement (so, to expand the example case above: [XPapyrus (Q#TBD)-->Location of discovery (P189)-->wall street, Dura-Europos (Q116621880)] [qualified by: Adjacent building (P3032)-->Block L7, Dura-Europos (Q108076746). So far, by my estimate, this is the solution that seems to best fit our use case; the only trouble is that the property right now specifies "building", and our use case would sometimes stipulate a building in the traditional sense, and other times would point to other kinds of archaeologically significant features like streets, city blocks, etc. Ahc84 (talk) 17:47, 2 June 2023 (UTC)[reply]

Perhaps @Jahl de Vautban, @Epìdosis may have thoughts on this, or know of others we should sound this out with? Ahc84 (talk) 17:53, 2 June 2023 (UTC)[reply]

The problem is interesting, but I think that as of now I don't have a good solution to propose, or a defined opinion. I'm interested in Jahl's opinion too. --Epìdosis 18:57, 2 June 2023 (UTC)[reply]

A few caveats before my main argument: data in general aren't really inaccuracy-friendly, and we are currently lacking any shapefile support to model lines or polygons. I'm not certain that we'll find a definitive answer. Anyway, some tought about it. For what I gather, you have two objectives: 1) modeling where a given find was found, with its inaccuracy; and 2) putting that on a map. Both objectives may be mutually incompatible.

2) seems the easiest to solve. To do that you'll need coordinate location (P625) (I think Pleiades also only works with coordinates). If you put them on the object of the location of discovery (P189) statement, you'll have a good many finds that share the same dot on the map. That can work if they were found in a single, indentifiable building, but on a street stretching the whole city that may not be what we want. So you'll probably end up putting it as qualifier under the object of the location of discovery (P189) to have a greater latitude (pun intended). As we can refine the coordinates to be more or less precise, with more or less digits, that may convey the inaccuracy of what a polygon could achieve.
This leads us to 1) and to what to put in location of discovery (P189) in the first place. You could use multiple location of discovery (P189), but, as you already feared, I expect any software to put several dots on the map and not to do a triangulation between a set of coordinates. This is what I meant when I said that 1) and 2) may be mutually incompatible, because the more you'll want to geographically describe the context of a find the more, paradoxically, you'll have difficulty in putting it on a map. This could be perhaps solved if the various statement share the same coordinate location (P625) qualifier and if we can somehow tell the software to discard duplicates, but I don't know if and how to do that. If placing the object on a map wasn't required I'd think this could be more accurate. As it's required however, I'd thus also think that a single location of discovery (P189) statement pointing to the least inaccurate location would be best, with qualifiers to further describe it, among which:
1. coordinate location (P625) as described above; alternatively that could be all we need and we just infer the proximity of a given feature by its coordinates;
2. adjacent building (P3032), while seemingly a good qualifier, has the caveat that Ahc84 described. We may want something more generic that would cover the case when something isn't an architectural structure;
3. shares border with (P47) could perhaps work in this regard, as more generic, but I don't know if we can use it for man-made topological features;
4. direction relative to location (P654) is another one, but I'm not sure that we can transform e.g. "found on X street, front of building Y" in "building Y, north of it" without losing the connection to the street, which seems more important; same goes with sourcing circumstances (P1480);
5. partially coincident with (P1382) that would describe the overlaps than you want, but I'm quite unsure on how to use it as qualifier;
6. significant place (P7153) doesn't seem suited to further describe a find, I gather it's more for the life ot the object (e.g. "Dura-Europos", "Yale", etc.);
7. location (P276) doesn't seem to be of much help as I think you'll need other qualifiers and we can't nest them;

Here what I can think of, not sure if that make us move very much forward. --Jahl de Vautban (talk) 07:17, 3 June 2023 (UTC)[reply]

Thanks so much for these suggestions Jahl! Some of them I've looked at, but not all. I added something to the discussion on adjacent building (P3032) where some people had already been discussing whether the scope of this property should be widened. Perhaps you might have a look there if you have a moment? I'd very much value your input. Property talk:P3032 -- JASHough (talk) 01:42, 6 June 2023 (UTC)[reply]

Wikidata talk:WikiProject Cultural heritage

Contents

To dos[edit]

Redundant[edit]

National Library of Wales[edit]

Notability and inclusion of replicable, non-artistic objects in heritage collections - input requested[edit]

COH Challenge[edit]

Mapping of South American datasets[edit]

Varying collection sizes for one institution[edit]

Copyright policy of an institution?[edit]

Feedback welcome: starting GLAM-related ontology and metadata mapping for Structured Data on Commons[edit]

Help with Lefkara lace[edit]

What are the best modelled items for your areas of interest?[edit]

Adding translations for the project pages[edit]

Describing location of buildings in archaeological sites[edit]

how to model inception for date ranges[edit]

Data modelling conventions of instance of (P31), genre (P136), levels of specificity, and consistency across cultural heritage description on Wikidata[edit]

"repository"[edit]

Name of Churches in Italian[edit]

Adding item Type Constraints for Historic Marker Properties[edit]

Poperty creation[edit]

Use of P527 vs P912 for parts of monuments[edit]

Proposal of update of format violations (P3503)[edit]

New project on Intangible Cultural Heritage[edit]

OPEN CALL FOR A WIKIPEDIAN IN RESIDENCE AT PAMM[edit]

Proposal for Cultural Good of Croatia ID[edit]

Question about instance of (P31)[edit]

Describing overlapping location data for find-spots[edit]

Navigation menu