Applying Topic Modeling for Automated Creation of Descriptive Metadata for Digital Collections

Creation of descriptive metadata for digital objects tends to be a laborious process. Specifically, subject analysis that seeks to classify the intellectual content of digitized documents typically requires considerable time and effort to determine subject headings that best represent the substance...

Full description

Saved in:
Bibliographic Details
Published inInformation technology and libraries Vol. 41; no. 2; pp. 1 - 14
Main Author Glowacka-Musial, Monika
Format Journal Article
LanguageEnglish
Published Chicago American Library Association 01.06.2022
Subjects
Algorithms
College Administration
Computational linguistics
Data mining
Digitization
Electronic documents
Electronic records
Humanities
Indexing
Information Services
Language Processing
Libraries
Library of Congress Subject Headings
Mathematics
Metadata
Modelling
Natural language interfaces
Natural language processing
Optimization
Primary Sources
Probability
Problem Solving
Subject heading schemes
Teaching Methods
Workflow
Online AccessGet full text
ISSN0730-9295
2163-5226
DOI10.6017/ital.v41i2.13799

Cover

More Information
Summary:Creation of descriptive metadata for digital objects tends to be a laborious process. Specifically, subject analysis that seeks to classify the intellectual content of digitized documents typically requires considerable time and effort to determine subject headings that best represent the substance of these documents. This project examines the use of topic modeling to streamline the workflow for assigning subject headings to the digital collection of New Mexico State University news releases issued between 1958 and 2020. The optimization of the workflow enables timely scholarly access to unique primary source documentation.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0730-9295
2163-5226
DOI:10.6017/ital.v41i2.13799