describing sentimental context of records

[nermeen ellaban]

Hi, Clavaud and everybody

Knowing how crucial it is to comprehend records, I'm curious about their context.
Lately, the sentimental context of textual records excites me. Because I believe that records, like humans, can express emotions disproportionately. I have analyzed sentiment in several Arabic, English, and Turkish records using machine learning applications on the web. I came to some significant conclusions that led me to propose that the new object properties be added to the standard "records in context(RIC)".

Object properties	value	IRI
rico:HasSentiment	Positive, negative, natural	https://www.ica.org/standards/RiC/ontology#HasSentiment
rico:HasEmotion	Fear, happy, sad, Satisfaction	https://www.ica.org/standards/RiC/ontology#HasEmotion

I sent you the RDF file, which includes the description of the Arabic record, its sentiment, and its emotion. (Object properties of sentiment and emotion are in red.)

I hope you find my suggestion intriguing.

I'm grateful to connect with you. I'm hoping to hear from you soon.

best regards 

Nermeen ellaban

assistant professor 

egypt

[nermeen ellaban]

---------- Forwarded message ---------
From: nermeen ellaban <nermeen...@gmail.com>
Date: Mon, Aug 5, 2024 at 5:25 PM
Subject: describing sentimental context of records
To: <Records_in_C...@googlegroups.com>

Hi, Clavaud and everybody

Knowing how crucial it is to comprehend records, I'm curious about their context.
Lately, the sentimental context of textual records excites me. Because I believe that records, like humans, can express emotions disproportionately. I have analyzed sentiment in several Arabic, English, and Turkish records using machine learning applications on the web. I came to some significant conclusions that led me to propose that the new object properties be added to the standard "records in context(RIC)".

Object properties	value	IRI
rico:HasSentiment	Positive, negative, natural	https://www.ica.org/standards/RiC/ontology#HasSentiment
rico:HasEmotion	Fear, happy, sad, Satisfaction	https://www.ica.org/standards/RiC/ontology#HasEmotion

I sent you the RDF file, which includes the description of the Arabic record, its sentiment, and its emotion. 

I hope you find my suggestion intriguing.

I'm grateful to connect with you. I'm hoping to hear from you soon.

[CLAVAUD Florence]

Dear Nermeen,

Thanks for sharing information on your project, and for contacting EGAD with a proposal!

Also, sorry for reacting a bit late. We are doing everything we can.

Indeed you are right. Archival records, and particulary items included in personal papers (textual documents like correspondence, diaries,  litterary manuscripts, and also graphic, photographic, audiovisual ones...) may express or evidence in various ways the sentiments or emotions of their authors and protagonists; and of course in turn the persons who access them may experience emotions - though it is not what you are talking about unless I am wrong.

Storing in archival metadata the emotions or sentiments expressed by such records is not classical nor frequent for now. The main reason being simply that most archivists usually do not have time for reading extensively the content of such items or analyzing them so precisely, having to deal with massive amounts of record sets, including the sets coming from public governments and agencies, where sentiments are less often expressed. When they have time, they usually just try to represent the topics of records, the things they are about, not what the authors felt.

So for now, RiC-O for now does not provide any solution to assert that a record expresses an emotion or sentiment. 

But today AI (and particularly the combination of HTR and NLP) can help to extract or discover many other features than topics from archival items, in addition to support classical metadata production.

Besides, researchers in social history (or others) may be interested in studying such features, as you did.

So, there may be research and development on such aspects, new trends and needs. More generally speaking, AI already has and will continue to  have significant impacts on archival metadata. RiC of course has to take these new perspectives and tools and into account.

We (I mean the RiC-O developement team) will discuss your proposal and see if we include it in the RiC-O roadmap for the coming months.

If we decide to add components to RiC-O to take this specific need into account, it may be in a very simple way for now, as you suggested, using a datatype property, so that archivists or researchers can easily assert that a rico:Record or a rico:RecordPart (as a sentence or phrase in a record can be considered a rico:RecordPart) expresses a feeling or an emotion.

Meanwhile, anybody can build an extension of RiC-O, and add to it what their project needs.

We would be interested in having more information and news from the project you mentioned.

And would also be glad to know if other user(s) there are interested in this topic.

I hope this answer suits you.

Best regards,

Florence Clavaud

(she/her/hers)
Executive member of ICA/EGAD ; lead of RiC-O development team
Conservatrice générale du patrimoine | General curator
Responsable du Lab des Archives nationales de France| head of the Lab, Archives nationales de France
Courriel : florence...@culture.gouv.fr

---

De : records_in_c...@googlegroups.com <records_in_c...@googlegroups.com> de la part de nermeen ellaban <nermeen...@gmail.com>
Envoyé : lundi 5 août 2024 16:27
À : Records_in_C...@googlegroups.com
Objet : [Records in Contexts users] Fwd: describing sentimental context of records

 

--
You received this message because you are subscribed to the Google Groups "Records_in_Contexts_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Records_in_Context...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/Records_in_Contexts_users/CAFpzGV%2B0z-%2Bz0Q4ed0TB%2BNr6Def%2BXKqOsATnQL8HYKf5jqsi-g%40mail.gmail.com.

---

Merci de nous aider à préserver l'environnement en n'imprimant ce courriel et les documents joints que si nécessaire.