Institutional Repository of Vilnius University
Propagandos ir dezinformacijos tekstų plitimo socialiniuose tinkluose tyrimas /
Abstract
dc:descriptionThe spread of propaganda and disinformation texts on social media is a relevant and complex contemporary issue significantly influenced by social media bots. This thesis analyzes the types of social bots and their detection methods, distinguishing the main groups of methods: machine learning, deep learning, graph-based, and expert-based methods. During the research, the level of bot prevalence on the social network X (Twitter) was evaluated using “Botometer X” across two propaganda-related and, for comparison, two non-propaganda hashtags. The results showed that bot prevalence under propaganda hashtags reached 13% and 8.7%, while under non-propaganda hashtags it reached 4.8% and 11.6%. When evaluating only currently active accounts (not deleted or suspended for violating platform rules), these indicators decreased. Applying the activity-based “intensive automation” method produced higher bot prevalence estimates, especially in propaganda-related hashtags. The calculated average bot prevalence level demonstrated that bot prevalence in propaganda hashtags was higher (18% and 14.1%) than in non propaganda hashtags (5.7% and 7.8%). The obtained results suggest that potentially automated accounts are used more frequently in propaganda-related discussions on social media, while different bot detection methods produce inconsistent results; therefore, a combination of multiple methods is necessary for more reliable evaluation.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Karvelis, Andrius,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD308337324&prefLang=en_US