Murtala Aminu Buhari
A Comparative Study on the Characteristics and Behaviours of Social Media Users in Response to Fake News on X(Twitter)
Buhari, Murtala Aminu; Mohammadi, Azadeh; Saraee, Mo; Olalekan, Salau Ayodeji
Authors
Dr Azadeh Mohammadi A.Mohammadi1@salford.ac.uk
Lecturer in Data Science
Prof Mo Saraee M.Saraee@salford.ac.uk
Interim Director of Computer Science
Salau Ayodeji Olalekan
Contributors
Murtala Buhari
Researcher
Dr Azadeh Mohammadi A.Mohammadi1@salford.ac.uk
Researcher
Prof Mo Saraee M.Saraee@salford.ac.uk
Researcher
Salau Ayodeji Olalekan
Researcher
Abstract
Social media has proven to be a game- changer in providing avenues for individuals from diverse back- grounds to communicate, exchange ideas, and enable friendships and businesses to reach new heights. Despite these profound advantages, it also has its downsides, especially when exposing users to fake news. The most worrisome aspect is when users react to such information using hate speech. Recent research has shown that fake news generates the most hate speech, as it evokes primordial sentiments of othering, dehumanization, and tribalism in multicultural societies. The study aims to compare the characteristics and behaviours of users engaged with hateful responses and those not engaged in hateful responses in code- mixed Nigerian English (broken) and English over events from 153 fake news posts from the political sphere in the Nigerian Election of 2023. The list of debunked fake stories was gathered from fact-checking websites, and the social media users who propagated these stories were identified. A corpus of 18,000 tweet replies was curated together with user profile information. A subset of 129 random hate speech and control non-hate speech user profiles was selected for this study. The main aim of the study was to investigate the unique characteristics of hate speech respondents and non-hate speech respondents by examining their profile descriptions. Hate users tend to have more followers, and engagement metrics than non- hate users even though they are not as active as non-hate users. Hate speech users had more blue verifications and location details on their profiles, but non- hate speech users’ locations had more identifiable real-world locations. The novel findings provide deeper insights into the characteristics of hate users, which can be used to develop features that will improve moderation and detection by machine learning algorithms.
Presentation Conference Type | Conference Paper (published) |
---|---|
Conference Name | 2024 IEEE 5th International Conference on Electro-Computing Technologies for Humanity (NIGERCON) |
Start Date | Nov 26, 2024 |
End Date | Nov 28, 2024 |
Acceptance Date | Oct 3, 2024 |
Online Publication Date | Mar 26, 2025 |
Publication Date | Nov 26, 2024 |
Deposit Date | Mar 30, 2025 |
Journal | 2024 IEEE 5th International Conference on Electro-Computing Technologies for Humanity (NIGERCON) |
Peer Reviewed | Peer Reviewed |
Pages | 1-5 |
DOI | https://doi.org/10.1109/nigercon62786.2024.10927148 |
Keywords | Measurement , Machine learning algorithms , Social networking (online) , Voting , Hate speech , Psychology , Speech recognition , Metadata , Linguistics , Fake news |
Publisher URL | https://ieeexplore.ieee.org/document/10927148 |
Make cities and human settlements inclusive, safe, resilient and sustainable
You might also like
Design of Named Entity Recognition Model for Awingi Language Using Deep Learning Approach
(2025)
Presentation / Conference Contribution
Performance Analysis of Different Shape Planner Inverted F Antenna for Unmanned Aerial Vehicle
(2025)
Presentation / Conference Contribution