User talk:Liber008

Welcome to Wikidata, Liber008!

Wikidata is a free knowledge base that you can edit! It can be read and edited by humans and machines alike and you can go to any item page now and add to this ever-growing database!

Need some help getting started? Here are some pages you can familiarize yourself with:

• Introduction – An introduction to the project.
• Wikidata tours – Interactive tutorials to show you how Wikidata works.
• Community portal – The portal for community members.
• User options – including the 'Babel' extension, to set your language preferences.
• Contents – The main help page for editing and using the site.
• Project chat – Discussions about the project.
• Tools – A collection of user-developed tools to allow for easier completion of some tasks.

Please remember to sign your messages on talk pages by typing four tildes (~~~~); this will automatically insert your username and the date.

If you have any questions, don't hesitate to ask on Project chat. If you want to try out editing, you can use the sandbox to try. Once again, welcome, and I hope you quickly feel comfortable here, and become an active editor for Wikidata.

Best regards! --Epìdosis 07:49, 21 October 2023 (UTC)[reply]

Bonjour et bienvenue sur Wikidata ! Je vois que tu as déjà trouvé le chemin des différents projets ayant trait à l'Antiquité, c'est parfait. N'hésite pas à poser des questions là-bas si tel ou tel point de Wikidata n'est pas clair ; tu peux aussi me contacter directement sur ma page de discussion ou bien Epìdosis au-dessus sur la sienne et on y répondra avec plaisir. Bonne continuation ! --Jahl de Vautban (talk) 17:59, 20 December 2023 (UTC)[reply]

Hi, I noticed that you added in some Greek items the instances of temple (Q44539) and archaeological site (Q839954), which are not needed when there's hieron (Q1617500) which is subclass of sanctuary, holy place and Ancient Greek archaeological site. Jimkats (talk) 15:52, 13 March 2024 (UTC)[reply]

Hi, thank you for your message. I added these specifically to do an easier request by SPARQL for greco-roman temples in Antiquity. As you know, temples are described can be described in a very heterogeneous way in Wikidata, so adding the more general instance of temple (Q44539) was for me a simple way to achieve my goal. But if there is a better way to do this, I can also just leave hiéron (Q1617500) and do a more complex query with SPARQL. Liber008 (talk) 16:04, 13 March 2024 (UTC)[reply]

Hi! First of all thanks for all the work you have been doing in the last months regarding classical antiquity in Wikidata, especially on the matching of Trismegistos god ID (P12467). I'm also a PhD in Classics (specifically Greek philology) and I have been active on Wikidata since 2013, trying (among many other tasks) to improve items regarding both ancient (and medieval) Greek and Latin authors and ancient mythology. I have recently started to work with @Anna Clara Maniero Azzolini: on a project of digital prosopography (User:Anna Clara Maniero Azzolini/Altinum) consisting in the import to Wikidata of items about the inscriptions from Altinum (Q441542) (first phase) and about the persons mentioned in these inscriptions (second phase). In the first phase of the work we analysed the terminology used by Epigraphic Database Roma (Q124541686), the source of our data regarding the inscriptions, and we improved or created the corresponding items in Wikidata, using as main reference point the EAGLE Vocabulary (Q19371183).

This evening I saw that you imported a relevant amount of data from the "inscription type" part of EAGLE Vocabulary to 103 Wikidata items; it is a very useful effort and I would like to thank you also for this task (BTW, I notice you have put "type of inscription in epigraphy" as description both in English and in French, probably you need to fix the French ones). However, I also need to notice that unfortunately, as I and Anna Clara have experienced in our work a few months ago, the EAGLE Thesaurus contains a lot of imprecise terminology which unfortunately cannot be safely used as labels and aliases in Wikidata and surely requires a manual check; here are a few examples of aliases I had to manually remove:

• sacred inscription (Q125323760) (e.g.): "inno sacro" doesn't mean this in Italian, and "religious" is evidently too generic in English
• honorary inscription (Q125323109) (e.g.): all these aliases are too generic or regard different concepts, like honorific statue (Q23830017)
• curse tablet (Q1430557) (e.g.): "tabellae defixionam" is a clear linguistic mistake
• Q126733424 (e.g.): "manumissione" is a linguistic mistake

I have also a few other doubts:

• I think Q126733452 and Q126733431 are not notable; I don't understand the meaning of Q126733484;
• subclass of (P279) is systematically used in the generic form subclass of (P279)inscription (Q1640824); this is often an overgeneralisation (see Wikidata:WikiProject Ontology/Issues#Overgeneralisation), since it would be better to use a more specific value, e.g. here, despite the fact that also EAGLE has no adequate ramification of subtypes;
• finally, in some cases I think EAGLE has not been reconciled correctly, e.g. international convention (Q30747863) (not all international conventions are inscriptions), cadastre (Q191072), constitution (Q7755), alphabet (Q837518) etc.; in general, since EAGLE entries regarding always concepts considered only as inscriptions, they deserve on Wikidata apposite items treating each concept only as inscription (e.g. "inscribed constitution", "inscribed international treaty"); saying e.g. that alphabet as a mathematic concept (alphabet (Q837518): non-empty set of symbols or letters that make up strings in a formal language) is a type of inscription and a subclass of inscription is surely incorrect.

In general, I would suggest a deep revision of all this import, mainly because the imported data contain themselves a wide range of mistakes and imprecisions, and also because in general managing the items of Wikidata regarding concepts is a very delicate task. If you have doubts about my comments, or in general if you want to discuss how to fix some specific problematic cases, I will be happy to discuss them with you. See you soon and thanks again for the work you are doing, --Epìdosis 20:34, 19 June 2024 (UTC)[reply]

Hi Epìdosis, thank you for your very important feedback since you are a reference in Wikidata's community! It's always a pleasure to chat with a fellow Ph.D. student and have some input. By the way, I planned to finish matching Trismegistos god ID (P12467) in the next months or so (with almost 3000 divinities, it is a lot of work!).

I just looked at your data model for inscriptions (User:Anna Clara Maniero Azzolini/Altinum) and it is a really good example how can epigraphy be imported in Wikidata more efficiently.

Regarding your comments about my upload yesterday of some of the EAGLE Vocabulary (Q19371183), I think your comments are spot on. I kinda took EAGLE authority a bit too seriously and just thought that their vocab would adequate for Wikidata. Looking back, there is indeed a lot of imprecision and generalisation on their part. With that in mind, I will:

• Fix the wrong description in French
• Try to spot aliases that are too generic/linguistically erroneus and remove them
• Ask for deletion international convention (Q30747863), Q126733431 and Q126733484 (which I don't really understand now that you mention it because EAGLE doesn't provide an example where this category is used
• Create distinctive elements for items such as alphabet (Q837518) or international convention (Q30747863) that, like you mentionned, are not fitting for inscription/type of inscription

I was initially planning to do other batches to import other type of inscriptions (as for now, I just imported them if EAGLE provided a French translation). What do you think I should do? Should I just stop for now? Would you be interested in mapping what could be imported in Wikidata and what shouldn't?

About the subclass of (P279)inscription (Q1640824), I knew this would eventually be a problem because indeed, it can be an overgeneralisation. I intended to initially do this import and after try to do ramification or subtype when all of the categories of inscriptions were imported in Wikidata.

Also, maybe this will be of interest for you, but I'm part of a research group (Wikidata:WikiProject_IDEA) that use Wikidata as a plateform to create Linked Open Data about Dura-Europos. We are currently working on importing Dura's inscriptions in Wikidata with a data model that is still a draft for now. Importing EAGLE vocab was part of this future import. I will get inspiration from your work on Altinum and see if our model matches yours. In any case, if you want to collaborate for future work on epigraphy, I would be happy to chat with you.

Thanks again for your suggestions and comments! Liber008 (talk) 14:04, 20 June 2024 (UTC)[reply]

Hi Liber008, thanks very much for your response. The matching of Trismegistos god ID (P12467) is a huge advance for Wikidata coverage of ancient mythology and religion, your work is really important. And you are the first PhD student in Classics active on Wikidata that I get to know (apart from me, obviously; but in fact my PhD project doesn't involve Wikidata), so surely I look forward to collaborating with you in the next months. My only limitations are time (which, unfortunately, is always very scarse for all the activities we do, both as classicists and as Wikimedians) and my area of specialization, which is somehow different: I am a hellenist and I especially study Greek rhetorical texts (cf. also this Wikibase instance about Greek rhetorical exercises that I have recently created) and my experience in the fields of ancient archaeology and epigraphy is unfortunately limited ... but I have at least a basic understanding of the related problems and I am glad to learn something more of course.

@Anna Clara Maniero Azzolini: is a very expert Latin epigraphist and is working on this very interesting project about the inscriptions from Altinum (Q441542) and the persons they mention; my contribution is mainly for the more technical aspects of Wikidata, but of course most of the merit for the reconciliation between EDR fields and Wikidata (User:Anna Clara Maniero Azzolini/Altinum) is hers.

Yeah, EAGLE Vocabulary has unfortunately an unsatisfying quality (I really hope that in the future someone will improve it; unfortunately it seems somehow stalled, all the entries I have consulted have been last edited many years ago), so it requires to be managed very carefully. Since it is an academic project, I perfectly understand that you just trusted it initially, so don't worry. All the four points of corrections that you outline are perfect in my opinion; for the deletion of the 3 items, I can do it directly as I am an administrator (BTW, the absence of examples and often also the absence of definition is one of the most disappointing flaws of the vocabulary, since its first scope as a controlled vocabulary would be exactly contributing to the construction of a rigorous terminology with proper definitions). My suggestion is to first clean through these four points this first part of the import and then proceed (carefully) with the other parts; yes, I could be interested in discussing what should be imported to Wikidata and what shouldn't - it won't be an easy discussion, since as I said working on terminology is usually very complex as we experienced also with Anna Clara, but it will surely be very interesting - let's talk about this via e-mail. I have some experience of controlled vocabularies because of my collaboration with the institution which manages the Italian national controlled vocabulary (see Wikidata:Gruppo Wikidata per Musei, Archivi e Biblioteche/Nuovo soggettario).

For subclass of (P279)inscription (Q1640824), no problem: your choice of first importing this generic statement and then fixing it manually afterwards is perfectly fine for me; I wrote this point yesterday just to be sure that you were aware of the problem, but solving it could be objectively difficult and it's not super-urgent, so take the necessary time to do it.

Fixing the terminology on Wikidata through an authoritative thesaurus before doing a massive import is of course a perfect course of work (the only problem was, as said, that EAGLE is not as authoritative as we would desire ...). Your data model for inscriptions (I suggest, when it is ready, to copy it from Google Drive also directly in the Wikidata page Wikidata:WikiProject IDEA/Inscriptions) is very complete and I really appreciate it. I have only one suggestion: having collection (P195) with qualifier inventory number (P217) is sufficient, adding also the inverse (i.e. inventory number (P217) with qualifier inventory number (P217)) is a redundancy and I would avoid it (see my essay Wikidata:Events/Data Quality Days 2022/Modeling data regarding issues like this). I see no relevant mismatches between your model and ours for Altinum; you may be interested in adding state of transmission (P12020) and height of letters (P12549); you could also have a look at Wikidata:Property proposal/writing technique and write your opinion, for the use of existing fabrication method (P2079) or for the creation of a new specific property. Yes, I would be interesting in collaborating with this project (although you should not expect to find in me an expert of epigraphy as I said); also Anna Clara could well be interested, and surely she is much more competent than me on the epigraphic part. Let's talk about this too via e-mail.

Of course for any doubt about fixes to EAGLE-related items, or other Wikidata doubts, feel free to write me here on Wikidata.

See you soon! Epìdosis 16:16, 20 June 2024 (UTC)[reply]

Salut, j'ai vu que sur Deiopea (Q59661541) tu as modifié la P31 de mythological Greek character (Q22988604) à Nereids (Q54230). J'avais justement fait l'inverse il y a trois mois. Avec le modélisation actuelle Nereids (Q54230)instance of (P31)group of Greek mythical characters (Q28061975), il me paraît ontologiquement très étrange qu'un personnage individuel comme Déiopé puisse être une instance d'un groupe et il me semblait plus logique d'indiquer cette relation avec part of (P361). La même problématique s'applique aux Okeanid (Q182406) (le libellé est au singulier, mais la P31 indique un groupe ce qui est un autre problème), aux Giants (Q191192), aux Muse (Q66016) et à tous les groupes de personnages en général. Qu'en penses-tu ? --Jahl de Vautban (talk) 13:53, 26 June 2024 (UTC)[reply]

Salut, merci pour le commentaire! Après réflexion, je crois qu'il serait effectivement étrange de mettre Néréides en instance of (P31) si c'est un groupe de personnage. Le pluriel est ici important considérant que c'est un groupe et non une seule Néréide. Il me semblait par contre flou et généraliste au premier coup d'oeil de mettre group of Greek mythical characters (Q28061975) en instance of (P31), mais part of (P361) peut rectifier le tir. Je suis donc d'accord qu'il faudrait plutôt le mettre en part of (P361).

Cela revient effectivement à faire le même processus pour tous les autres types de groupe de la mythologie grecque comme Giants (Q191192) ou Muse (Q66016). Il faudrait éventuellement faire un ménage donc pour tous ces éléments.

Je note d'ailleurs qu'il existe deux instances pour Océanide, une au singulier Okeanid (Q182406) et une au pluriel Oceanids (Q20575671). Ces deux éléments seraient susceptibles pour un fusion, je n'arrive pas à déceler de différence. Liber008 (talk) 14:55, 26 June 2024 (UTC)[reply]

Une proposition concernant les différents groupes de nymphes grecques serait de mettre en instance of (P31) la valeur Greek nymph (Q3346693) qui est à mon sens plus précise que mythological Greek character (Q22988604). Qu'en penses-tu? Liber008 (talk) 17:19, 26 June 2024 (UTC)[reply]

Greek nymph (Q3346693) convient tout à fait, via Greek deity (Q22989102) c'est une sous-classe de mythological Greek character (Q22988604) et c'est effectivement plus précis. J'ai jeté un œil aux différentes P31 des éléments liés à la mythologie grecque, sans surprise c'est un peu le bazar avec plus de 300 P31 différentes. Je vais essayer d'en normaliser une partie et d'en tout cas résoudre les groupes.

La petite subtilité pour Okeanid (Q182406) et Oceanids (Q20575671), c'est que le premier élément est limité aux filles d'Océan, tandis que le second est pour les filles et les fils (Potamoi (Q3027575)). En pratique, j'ai surtout l'impression que la distinction n'existe qu'à cause des deux articles sur Wikipédia en breton et le Qid pour Oceanids (Q20575671) ne me paraît pas vraiment nécessaire. @VIGNERON: pourrais-tu regarder s'il est possible de fusionner ces deux articles ? --Jahl de Vautban (talk) 18:31, 26 June 2024 (UTC)[reply]

Concernant notre problème @Jahl de Vautban, je viens de remarquer que Infovarius vient d'annuler une de mes modifications en remettant naiad (Q182037) en instance of (P31) plutôt qu'en part of (P361) avec comme commentaire: "rather a class (better than a definite group". Liber008 (talk) 13:51, 30 June 2024 (UTC)[reply]

@Infovarius: not sure if you're able to follow what precedes, but in short Greek divinites are at the moment badly modeled, with various concepts defined as group being used as P31 for single characters. naiad (Q182037) is one such exemple, Oneiroi (Q578584) is an other, etc. So either we change the modelling of some concepts, or we need another property than P31 to reflect their link to the concept. --Jahl de Vautban (talk) 17:28, 1 July 2024 (UTC)[reply]