Identifying the Impact of Propagandistic Social Media Accounts
The methodology will combine state-of-the-art Artificial Intelligence methods with investigative journalistic tactics.
Digital platforms, especially social media sites, have become leading channels for public discourse. Users propagate their ideologies and beliefs without too much fear of official censorship. However, adversaries of such propagandist posts are beginning to work at a subtler level to impose censorship. A section of this adversarial group uses fake news, identities, or profiles to target propagandist posts and inflame communal emotions in society.
This project aims to develop a comprehensive methodology to:
- Identify fake online social network accounts.
- Map their connections to other accounts.
- Monitor and assess their activities.
The methodology will combine state-of-the-art Artificial Intelligence methods with investigative journalistic tactics, using the case study of Kyrgyzstan. We have chosen Kyrgyzstan because it has a much more free media environment than all its neighbors with very little official censorship or surveillance. Conversely, self-censorship is well evident in the local journalistic community.
By automating detection of fake accounts, the project aims to analyze the extent, tactics, and impact of the effort. We will adopt an interdisciplinary approach to the problem, combining machine learning, philosophy, and journalism. We will take cognizance of interview studies, manual data collection and reporting from locals, and automated data analysis.
The study could help social media platforms combat the presence and influence of fake identities or accounts.