Wikidata:Property proposal/CDDA designationType
designation type
[edit]Originally proposed at Wikidata:Property proposal/Natural science
Motivación
[edit]The Nationally designated areas inventory (Q1116062) includes lots of details of official protected areas in the European Environment Information and Observation Network (Q1378198). It's hard to characterize a protected area (Q473972) since there are dozens of varieties, a lot of more if you consider all the designations each country uses. These not always can be mapped easily so the EIONET chose to identify each national designation and assigns a unique ID with the designationTypeCode attribute. The European Environment Information and Observation Network (Q1378198) database (more than 60K items) characterize each entry with the designationTypeCode and the IUCN protected areas category (P814). With the proposed property we can use that code for any CDDA included entry. Through the property we can link the protected area with related public office and law references.
This proposal complements with Wikidata:Property proposal/CDDA designationTypeCode.
—Ismael Olea (talk) 22:57, 9 August 2021 (UTC)
Discussion
[edit]- Support ArthurPSmith (talk) 16:49, 10 August 2021 (UTC)
- Comment if Wikidata:Property proposal/CDDA designationTypeCode maps national designation types, shouldn't the values on protect areas be national designations and the label of the proposed property reflect that? BTW The proposed datatype is "item", but the samples currently use string. Maybe it's clearer once we see three samples with actual items as values. You could reuse the ones made for the other proposal. --- Jura 10:30, 11 August 2021 (UTC)
- @Jura1 Yes, I made a system to use the national designation as main label in native language and as official name (P1448) and the English name is used in the EN label. This is easy to do because data is provided in the source database. You can check some elements I'm uploading. I've updated the examples in the proposal as you suggest too. —Ismael Olea (talk) 14:42, 11 August 2021 (UTC)
- It's much clearer now with the updated samples. Thanks! Couldn't we label this property simply "protected area designation"? This would be consistent with heritage designation (P1435). --- Jura 10:28, 12 August 2021 (UTC)
- @Jura1 I'm not sure about a more generic labeling because this CDDA database covers countries in the European Environment Information and Observation Network (Q1378198), European and Euroasain, AFAIK :-/ —Ismael Olea (talk) 11:03, 12 August 2021 (UTC)
- I don't see a downside if the property also includes classifications for countries that aren't mapped by that database. --- Jura 11:11, 12 August 2021 (UTC)
- Honestly, it's not a thing about downsides but I'm simply not sure if it would be practical or not. Don't have experience neither confident enough to evaluate if would imply a potential future misunderstandings or not. This is why I act conservative. OTOH, if you and others consider the change appropriate, by the same premise I can't be against :-) —Ismael Olea (talk) 11:49, 12 August 2021 (UTC)
- It's much clearer now with the updated samples. Thanks! Couldn't we label this property simply "protected area designation"? This would be consistent with heritage designation (P1435). --- Jura 10:28, 12 August 2021 (UTC)
- @Jura1 Yes, I made a system to use the national designation as main label in native language and as official name (P1448) and the English name is used in the EN label. This is easy to do because data is provided in the source database. You can check some elements I'm uploading. I've updated the examples in the proposal as you suggest too. —Ismael Olea (talk) 14:42, 11 August 2021 (UTC)
- Support User:JavierMunozF (talk) 22:33, 11 August 2021 (UTC)
- Comment Given the above discussion I wonder if it would be simpler to just rely on instance of (P31) for this purpose? ArthurPSmith (talk) 17:20, 12 August 2021 (UTC)
- Comment Some premises seem off here. As slightly covered in above discussion, designations in question are not of or by CDDA dataset. This dataset just references and records national designations that for the most part existed in other (national) databases before this dataset was even created. So it's misleading to say these are "CDDA designation types" or alike. For the same reason it also looks odd to set these designations as instances of designation type (Q108028209) [update: as of August 19th this metaclass has been amended]. Similarly, in related database there are also IDs for species (EUNIS ID for species (P6177)), but we probably wouldn't classify species with EUNIS ID as "EUNIS species". Possibly items for Spanish national designations (examples above) didn't exist on Wikidata before, but several others are the very same that have been in use for a while, e.g. Nationalpark (Q21815132) (CDDA code DE01) or nature reserve in the Czech Republic (Q21101734) (CDDA code CZ02).
instance of (P31), or heritage designation (P1435) in case of serveral countries, is already used to link to these designations. If we seek consistency, then I think there wouldn't be any significant downside if existing heritage designation (P1435) was used for all such designations. The problem with something like "protected area designation" would be that it would contribute to an endless confusion around the fact not all protected natural objects are protected areas. CDDA dataset, WDPA and other related databases also include designations for protected natural objects that are not protected areas (like variants of "natural monument" designation for individual boulders and trees). For such objects designation being given as P31 also doesn't seem quite right as individual tree or boulder isn't essentially its current protection status (designation). Previously I've commented on this issue more thoroughly e.g. here. 2001:7D0:81DA:F780:D0F2:C01B:737C:BC2B 08:54, 15 August 2021 (UTC)
- Also description for designation type (Q108028209) ("developed by the CDDA to classify protected areas according national regulation") is simply wrong. Designations in question are developed/assigned by national authorities, and CDDA (EIONET) just references and uses these as they are. The same way it's odd to define individual designations like nature reserve (Q108061047) (duplicating Q28055269) by association to CDDA, while CDDA is merely one of several datasets where national designations can be found. 2001:7D0:81DA:F780:8CA:CAD8:3CA5:E487 17:30, 15 August 2021 (UTC)
- The duplicity of designations, like the case you explain is because it's almost impossible to me the identify en WD the correct ones. Also, it could depend on how wikidaters organize their country data (you know there are more or less sutiles differences). The good thing is the problem is really easy to solve just merging with the correct ones by someone with knowledge about. This is why I added the legal references of each case. I know is not a perfect job but it's a beginning I expect to be useful. Personally I made some effort to detect correct elements but for the most of the languages is near to imposible to me to do without help.
- About the labeling/descripting problem, would be great to hear your suggestions. Seems you have better knowledge than me and I'm always open to enhancements 👍 —Ismael Olea (talk) 19:08, 15 August 2021 (UTC)
- My suggestion is to drop "CDDA" from propety label, and otherwise not associate given designation to CDDA apart from just providing CDDA designation code among other identifiers. Or rather consider this property proposal redundant to already existing heritage designation (P1435). Above some other more specific label than that of P1435 has been considered, related to natural heritage, but note that CDDA data anyway also references designations that are not for natural heritage, such as code EE25 for cultural heritage. 2001:7D0:81DA:F780:4CCD:3CA1:A616:3488 09:23, 16 August 2021 (UTC)
- Also description for designation type (Q108028209) ("developed by the CDDA to classify protected areas according national regulation") is simply wrong. Designations in question are developed/assigned by national authorities, and CDDA (EIONET) just references and uses these as they are. The same way it's odd to define individual designations like nature reserve (Q108061047) (duplicating Q28055269) by association to CDDA, while CDDA is merely one of several datasets where national designations can be found. 2001:7D0:81DA:F780:8CA:CAD8:3CA5:E487 17:30, 15 August 2021 (UTC)
- I too have mixed feelings about this property, but I see it as similar with IUCN protected areas category (P814) with the main difference that there is usually a 1:1 correspondence between the national protected area type and the ID, so it is somewhat redundand information. But same applies to IUCN category, at least in my experience all items of a given national type have the same IUCN category as well. Extended this property to be heritage designation (P1435) for natural sites IMHO would be a can of worms, as it will make it even easier to declare a geographic feature being the same as a protected area, or putting together different protected areas in one item. Just look at the constraint violations for WDPA/CDDA/Natura2000-ID to see how many wrongly modeled protected areas we have in our database already. Ahoerstemeier (talk) 21:57, 15 August 2021 (UTC)
- @Ahoerstemeier I agree it's a nightmare to describe protected areas in a systematic way in WD. The good thing of the WDPA/CDDA/Natura2000 id's is considering we can't faithfully describe the elements at least we can reduce the ambiguity a bit identifing each area with (more or less) international way. Plus, the CDDA type codes adds a bit of description (something more than IUCN code). The challenge would be to establish a worldwide ontology but it's absolutely out of my current goals. —Ismael Olea (talk) 08:30, 16 August 2021 (UTC)
- I don't think that the use of heritage designation (P1435), or seprate "natural heritage designation" property, would make it easier to mix protected areas and other distinct geographical objects. In most cases where this has happened designations are currently given as P31 values anyway. Also please note that, as pointed out above, WDPA/CDDA include many individual protected objects that are *not* protected areas, and so many of these WDPA/CDDA property constraint violations are in fact false positives, e.g. in case of protected object Labidakivi (Q12368168). Considering this, as far as I can see and as previously explained e.g. here, the use of heritage designation (P1435) would rather help clear things up.
- As for IUCN categories, it's true that protected objects of some designations fall into the same IUCN category, but there are also other designations for which it isn't true. For instance, per CDDA data, objects of designation code EE12 are divided between IUCN categories Ib, IV and V. 2001:7D0:81DA:F780:4CCD:3CA1:A616:3488 09:23, 16 August 2021 (UTC)
- Maybe adding a restriction in WDPA/CDDA/Natura2000 id properties to be used in P1435 instead of P31 would help to reduce the mess? Or viceversa, whatever gets agreed. I can't have a preference for neither one approach. —Ismael Olea (talk) 10:00, 16 August 2021 (UTC)
- I too have mixed feelings about this property, but I see it as similar with IUCN protected areas category (P814) with the main difference that there is usually a 1:1 correspondence between the national protected area type and the ID, so it is somewhat redundand information. But same applies to IUCN category, at least in my experience all items of a given national type have the same IUCN category as well. Extended this property to be heritage designation (P1435) for natural sites IMHO would be a can of worms, as it will make it even easier to declare a geographic feature being the same as a protected area, or putting together different protected areas in one item. Just look at the constraint violations for WDPA/CDDA/Natura2000-ID to see how many wrongly modeled protected areas we have in our database already. Ahoerstemeier (talk) 21:57, 15 August 2021 (UTC)
- Support - Salgo60 (talk) 10:24, 15 August 2021 (UTC)
- Support --Tinker Bell ★ ♥ 18:46, 15 August 2021 (UTC)
- Comment I don't think this proposal is quite ready yet, and needs more thought . The primary reason is that the EIONET designation-types schemas might still be in a non-ratified or draft form as seen at https://dd.eionet.europa.eu/schemasets/browse/ and where EIONET has not updated to a ratified state of it's own documentation. (someone should push them to that stage before we make decisions on Wikidata regarding Linked Data. 2. The vocabulary and base URI's are documented, and where there is not just 1 "designation type" as this proposal protrays, but 2 identifier types: A designationTypeCategory https://dd.eionet.europa.eu/vocabulary/cdda/designationTypeCategory/view and designationTypeCodeValue https://dd.eionet.europa.eu/vocabulary/cdda/designationTypeCodeValue/view and finally where the RDF is available from their link here https://dd.eionet.europa.eu/vocabularyfolder/cdda/rdf
But more importantly, the Definition of the Nationally Designated Areas does include some verbiage about the designationTypeCategory aligned with IUCN, but lacks clarity mentioning mapping for the Codes, which can be seen if scrolled down on https://dd.eionet.europa.eu/vocabulary/cdda/designationTypeCodeValue/view
The CDDA Codes however could certainly be mapped to Wikidata items of the exact same concept, just as IUCN categories are now linked to Wikidata items, Ex. IUCN category II: National Park (Q14545628) which is different from the more general concept of national park (Q46169) and were the CDDA Codes could be used as External Identifiers on either the Wikidata general concepts like national park (Q46169) where many are instances of administrative territorial entity type (Q15617994), or mapped to the Wikidata items that are instances of IUCN management category (Q3679744) , by either perhaps were applicable and it makes sense. There certainly might be some overlap but we can always use different from (P1889) or other appropriate properties.
So I'd like to see 3 things on this proposal before supporting it.- 1. Update the name to the more appropriate "CDDA Designation Type Code" which represents the Code's Value as shown.
- 2. Update the examples to use their identifier. Ex. AZ04 = Ecological Park
- 3. Change the data type to be an External Identifier that takes a format that would allow URL syntax such as http://dd.eionet.europa.eu/vocabulary/cdda/designationTypeCodeValue/AZ04 -- Thadguidry (talk) 02:08, 16 August 2021 (UTC)
- @Thadguidry Did you checked the complementary proposal Wikidata:Property proposal/CDDA designationTypeCode? I think it answers at least to some of your suggestions. —Ismael Olea (talk) 08:20, 16 August 2021 (UTC)
- @—Ismael Olea Ah! Yes, then it makes much more sense. And your example links above indeed point to distinct entities representing a named/labeled/identified protected zone in a jurisdiction which means you can then say lots of things about that distinct protected zone Wikidata item such as when it was created, which laws, external id's, who maintains or operates it, etc. What then happens is that you have created a new Wikidata item that represents the same concept as an external entity. natural park (Q107690235) is exact match (P2888) to http://dd.eionet.europa.eu/vocabulary/cdda/designationTypeCodeValue/ES10. But this is fine, sometimes we do this alot, where we don't just simply link, but copy an entire external concept and most of it's data into a new Wikidata item to further use, link with, and add additional statements to. In my world, this is https://en.wikipedia.org/wiki/Data_fusion and a very important role of Wikidata to slowly copy and hold the world's structured data. Anyways, I'll change my vote now. -- Thadguidry (talk) 12:37, 16 August 2021 (UTC)
Support Thadguidry (talk) 12:37, 16 August 2021 (UTC)
- Oppose prposed label (see comment above for alternative label). --- Jura 16:30, 16 August 2021 (UTC)
- @Jura1 I removed the CDDA label and uptated the CDDA designationTypeCode (P9800) property mentions. —Ismael Olea (talk) 17:18, 19 August 2021 (UTC)
- I find it less confusing without CDDA associated label, and domain not being limited to protected areas makes sense as CDDA as a source includes also designations for individual natural objects that are not areas, as well as cultural heritage designations (references above). But now, with "designation type" label, it's pretty much the same as already existing heritage designation (P1435). 2001:7D0:81DA:F780:ED98:52B1:1627:75F2 07:52, 20 August 2021 (UTC)
- Yeah, somehow the label should specify what the designations apply to and differentiate from P1435. --- Jura 13:15, 20 August 2021 (UTC)
- Or, simply the existing P1435 property could be used instead of this proposed property. As pointed out above, it is already used for several natural heritage designations, and above I don't see a clear reason why it shouldn't be. 2001:7D0:81C5:A580:708F:5A45:F175:EC01 09:16, 7 April 2022 (UTC)
- Withdrawn And thanks all for your feedback. —Ismael Olea (talk) 08:25, 29 August 2023 (UTC)