#NoMore: An Analysis of Topics and Sentiments Indicative of the War in Ethiopia

Zeleke, Meseret; Hong, Lingzi; Kaz-Onyeakazi, Ijay; Smith, Daniella; Milburn, Stacie; Esener, Yildiz

Permalink

https://hdl.handle.net/2142/117377 Copy

Description

Title

#NoMore: An Analysis of Topics and Sentiments Indicative of the War in Ethiopia

Author(s)

• Zeleke, Meseret
• Hong, Lingzi
• Kaz-Onyeakazi, Ijay
• Smith, Daniella
• Milburn, Stacie
• Esener, Yildiz

Issue Date

2023-03-13

Keyword(s)

• Twitter sentiment analysis
• Ethiopia
• #NoMore
• Latent dirichlet allocation (LDA) and Hashtag activism

Abstract

Understanding public opinion about a conflict is essential in building democracy, cooperative relationships, trust, and protecting human rights. This research dissects public opinions on Twitter concerning the conflict in Ethiopia. This study utilized 707,720 public tweets with the #NoMore hashtag for analysis using Latent Dirichlet Allocation (LDA), and six dominant discussion topics were identified. Each topic represented at least 12.6% of the tweets, indicating that each topic was well supported in conversation. A pre-trained machine learning model was applied to evaluate the sentiments of each tweet. The analysis showed that many Twitter users believed mainstream media, international organizations, and countries external to Ethiopia might be negatively influencing the conflict and that tweets were predominantly objective and had neutral polarity. Past studies refer to similar Twitter dialogues as “hashtag activism,”, which uses hashtags to establish support and bring real-world changes through social media. Our in-depth public opinion analysis will provide insights to improve various parties' existing engagements, policies, laws, and decision-making processes for the benefit of the community at large.

Publisher

iSchools

Series/Report Name or Number

iConference 2023 Proceedings

Type of Resource

• Other
• text

Language

eng

Handle URL

https://hdl.handle.net/2142/117377

Copyright and License Information

Copyright 2023 is held by Meseret Zeleke, Lingzi Hong, Ijay Kaz-Onyeakazi, Daniella Smith, Stacie Milburn, and Yildiz Esener. Copyright permissions, when appropriate, must be obtained directly from the authors.

Owning Collections