Periods on Paper Topic Modeling and Sentiment Analysis of Menstruation in 20th-Century Dutch Newspapers
Keywords
Loading...
Authors
Issue Date
2025-08-01
Language
en
Document type
Journal Title
Journal ISSN
Volume Title
Publisher
Title
ISSN
Volume
Issue
Startpage
Endpage
DOI
Abstract
Menstruation remains a stigmatized topic in many societies. Newspapers shape and reflect public
opinion, yet there is little research on the coverage of menstruation in Dutch newspapers. This thesis
addresses that gap by using topic modeling and sentiment analysis to explore how menstruation was
discussed in the Netherlands throughout the twentieth century. Using menstrual-related keywords,
9712 relevant articles were retrieved from the Delpher newspaper archive, and the influence of OCR
errors and the disambiguation of keywords was assessed. With Latent Dirichlet Allocation (LDA),
seventeen distinct topics were identified, including abortion, contraceptives, menopause, religion,
and women’s health. The sentiment analysis, using a fine-tuned version of RobBERT, revealed 7419
neutral, 1939 positive, and 354 negative newspaper articles. The ideological background (pillarization)
of newspapers influenced how menstruation was covered, both in terms of topic and sentiment.
However, whether a newspaper was national or local had minimal impact. This research highlights
how computational methods can uncover historical patterns in sensitive or stigmatized topics.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen