Digital repository of Slovenian research organisations

Show document
A+ | A- | Help | SLO | ENG

Title:Umetna inteligenca kot ključni dejavnik pri strokovni obdelavi avdiovizualnega dokumentarnega in arhivskega gradiva
Authors:ID Dornik, Boštjan (Author)
Files:URL URL - Source URL, visit https://www.pokarh-mb.si/storage/app/media/Moderna_arhivistika_2025_1/01_Dornik_2025.pdf
 
.pdf PDF - Presentation file, download (1,04 MB)
MD5: 4F68F02A4C3238DA7FB3CBC24E1A8A40
 
Language:Slovenian
Typology:1.01 - Original Scientific Article
Organization:Logo PAM - Regional Archives Maribor
Abstract:Namen raziskave je bil raziskati vlogo umetne inteligence pri strokovni obdelavi avdiovizualnega arhivskega gradiva, s posebnim poudarkom na optimizaciji popisovanja in izboljšanju dostopnosti vsebin. V prispevku želimo prikazati, kako umetna inteligenca rešuje izzive, kot so obsežnost gradiva, časovna zamudnost ter kadrovska podhranjenost. Raziskava je temeljila na kombinaciji kvalitativnih metod, ki vključujejo pregled mednarodne literature o uporabi umetne inteligence v arhivistiki ter študijo primera na Radioteleviziji Slovenija, kjer so testirali integracijo umetne inteligence v arhivski sistem Mediateke. Poseben poudarek je bil namenjen analizi generativnih modelov, prepoznavanju govora, analizi sentimenta in razvrščanju vsebin. Rezultati raziskave kažejo, da umetna inteligenca omogoča hitrejšo in metapodatkovno bogatejšo obdelavo gradiva (npr. prepoznava tem, entitet, žanrov in sentimenta). Primerjava ročnega in avtomatiziranega popisa je pokazala bistveno zmanjšanje potrebnega časa (iz ur v minute) in večjo konsistentnost metapodatkov. Pri obdelavi avdiovizualnega gradiva se še pojavljajo slovnične napake in druge omejitve, zato je potrebna nadaljnja optimizacija delovnih tokov in integracija s sistemi za nadzor kakovosti. Na podlagi ugotovitev lahko zaključimo, da je umetna inteligenca ključna za učinkovito upravljanje vse obsežnejših digitalnih arhivov. Najobetavnejše funkcionalnosti vključujejo zvočno transkripcijo in prepoznavanje imenovanih entitet, ki se bosta v prihodnosti nadgradila z optičnim prepoznavanjem znakov (OCR) ter z orodji za prepoznavo slikovnega gradiva. Odprto ostaja tudi vprašanje, ali naj arhivi razvijejo nove metapodatkovne strukture, ki jih omogoča umetna inteligenca, ali pa naj se osredotočijo na prilagajanje obstoječih standardov.
Keywords:avdiovizualno gradivo, televizijski arhivi, strokovna obdelava, umetna inteligenca, sistemi za avtomatiziran popis
Publication date:01.01.2025
Year of publishing:2025
Number of pages:str. 1-14
Numbering:Letn. 8, št. 1
PID:20.500.12556/DiRROS-31535 New window
UDC:930.25:004.9
ISSN on article:2591-0884
DOI:10.54356/MA/2025/RLYU9154 New window
COBISS.SI-ID:250551043 New window
Note:Nasl. z nasl. zaslona; Opis vira z dne 26. 9. 2025;
Pub. date in DiRROS:05.08.2026
Views:241
Downloads:134
Metadata:XML DC-XML DC-RDF
:
Copy citation
  
Share:Bookmark and Share



Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Record is a part of a journal

Title:Moderna arhivistika : časopis arhivske teorije in prakse
Publisher:Pokrajinski arhiv
ISSN:2591-0884
COBISS.SI-ID:292560384 New window

Secondary language

Language:English
Title:Artificial intelligence as a key factor in the professional processing of audiovisual current and archival records
Abstract:The purpose of this research was to investigate the role of artificial intelligence in the professional processing of audiovisual archival material, with a particular focus on the optimisation of description and improvement of content accessibility. The article aims to demonstrate how artificial intelligence addresses key challenges such as the vast volume of material, time-consuming manual processing, and staff shortages. The research employed a combination of qualitative methods, including a review of international literature on the application of artificial intelligence in archival science, as well as a case study at Radiotelevizija Slovenija, where the integration of artificial intelligence into the Mediateka archival system was tested. Special attention was given to the analysis of generative models, speech recognition, sentiment analysis, and content classification. The findings indicate that artificial intelligence enables faster and metadata-rich processing of material (e.g., recognition of topics, entities, genres, and sentiment). A comparison between manual and automated description revealed a significant reduction in processing time (from hours to minutes) and greater metadata consistency. However, challenges remain, particularly in terms of grammatical errors and other limitations in the processing of audiovisual content, underscoring the need for further workflow optimization and integration with quality control systems. Based on the results, it can be concluded that artificial intelligence is essential for the effective management of increasingly extensive digital archives. The most promising functionalities include audio transcription and named entity recognition, which are expected to be enhanced in the future with optical character recognition (OCR) and image recognition tools. An open question remains whether archives should develop new metadata structures enabled by artificial intelligence or focus on adapting existing standards.
Keywords:audiovisual materials, television archives, professional processing, artificial inteligence, automated description systems


Back