Digital repository of Slovenian research organisations

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

Title:Ekstrakcija metapodatkov s pomočjo strojnega učenja
Authors:ID Sabadin, Ivančica (Author)
Files:URL URL - Source URL, visit http://www.pokarh-mb.si/storage/app/media/Moderna_arhivistika_2023_2/06%20Sabadin_2023.pdf
 
.pdf PDF - Presentation file, download (924,52 KB)
MD5: F66FB5D6574D67B36714E8970BB54013
 
Language:Slovenian
Typology:1.02 - Review Article
Organization:Logo PAM - Regional Archives Maribor
Abstract:Namen prispevka je raziskovati tehnike ekstrakcije metapodatkov s pomočjo strojnega učenja. Uporabljena je bila metoda pregleda literature iz podatkovnih baz ProQuest, Scopus in Emerald Insight. Rezultati so pokazali, da so tehnike strojnega učenja že uveljavljene na področju ekstrakcije metapodatkov iz znanstvene literature. Najboljše rezultate so pokazale rešitve, ki združujejo analizo postavitve dokumenta in proces ekstrakcije metapodatkov. Glede na raziskave o ekstrakciji metapodatkov s pomočjo strojnega učenja lahko sklepamo, da je treba dodatno analizirati orodja in modele strojnega učenja GROBID, CERMINE, XTRACT, BERT, Mask R-CNN in BiLSTM. Na podlagi izkušnjah ekstrakcije metapodatkov iz znanstvene literature je treba prilagoditi modele za ekstrakcijo metapodatkov iz arhivskega gradiva.
Keywords:ekstrakcija metapodatkov, strojno učenje, nadzorovano učenje, modeli strojnega učenja, obdelava naravnega jezika
Publication status:Published
Publication version:Version of Record
Publication date:01.01.2023
Year of publishing:2023
Number of pages:str. 255-269
Numbering:Letn. 6, št. 2
PID:20.500.12556/DiRROS-31618 New window
UDC:930.25:004.9
ISSN on article:2591-0884
DOI:10.54356/MA/2023/VRNY7665 New window
COBISS.SI-ID:181337859 New window
Note:Nasl. z nasl. zaslona; Opis vira z dne 3. 1. 2024;
Publication date in DiRROS:05.08.2026
Views:36
Downloads:20
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:Metadata extraction using machine learning
Abstract:The aim of this paper is to explore metadata extraction techniques using machine learning. The method used was a literature review of the ProQuest, Scopus and Emerald Insight databases. The results showed that machine learning techniques are already well established in the field of metadata extraction from scientific literature. The best results were shown by solutions that combined document layout analysis and metadata extraction processes. Based on the research on metadata extraction using machine learning, it can be concluded that further analysis of the machine learning tools and models GROBID, CERMINE, XTRACT, BERT, Mask R-CNN and BiLSTM is needed. Based on the experience with metadata extraction from scientific literature, the models should be adapted for metadata extraction from archival material.
Keywords:metadata extraction, machine learning, supervised learning models, machine learning model, natural language processing


Back