Twitter Sentiment Analysis of the 2020 United States Democratic Party Presidential Primaries
Keywords
No Thumbnail Available
Authors
Issue Date
2020-02-04
Language
en
Document type
Journal Title
Journal ISSN
Volume Title
Publisher
Title
ISSN
Volume
Issue
Startpage
Endpage
DOI
Abstract
Social media provide a lot of information. Processing this data manually is a laborious and cumbersome task. Sentiment Analysis could speed up this process significantly. The goal of this research was to find out whether Sentiment Analysis could be a valid tool for the prediction of the public opinion with respect to the 2020 Democratic Party Presidential Primaries. The data for the research was gathered on Twitter. Pre-processing was done in multiple stages in which Twitter-specific symbols were removed and the tweets were sorted per candidate per debate. Multiple sentiment analysis models, both machine learning and lexicon based, were tested and compared after which a Support Vector Machine proved to be the most capable model under the circumstances. The gathered data was then analysed using this model. The results provided insufficient proof for the hypotheses to be accepted but did establish a foundation for future research.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen