DOSSMA – Detection of Suspicious Social Media Activities

Description

The DOSSMA project will investigate suspicious and malicious behaviour on social media platforms. In a first phase, we will compile an extensive survey report on the areas that are currently being researched, including the respective state-of-the-art, existing solutions and initiatives. This report will serve as a basis to pinpoint one or two suited topics where our expertise in the NLP domain is a good fit, and we will aim to make a scientific contribution to advance the state-of-the-art in a second phase of the project. In preliminary work, we have identified the following topics and questions to be relevant and of interest for the survey:

Hijacked accounts: How can hijacked accounts be identified automatically if there is no filed complaint by the original user?
Impersonation: How can accounts of users pretending to be someone else be identified and distinguished from the actual account of the impersonated user? Can false demographic information (such as incorrect age information) be uncovered from properties of language use?
Malicious content: What are current issues and roadblocks in the automatic detection of Propaganda, Toxic Discourse, Hate Speech, Harassment, Conspiracy Theories, and Racism on social media?
Astro-Turfing, Fake Movements, and False Flag Operations: How can social media activities be recognized that are deliberately framed as being initiated by someone else?
Fake News: How can the posting and propagation of fake news be recognised and what are current efforts of e.g. large companies?
Radicalization: Can techniques for detecting malicious content be applied to uncover networks of radicalization?
Bot Recognition: How can a bot account or bot networks aimed to propagate malicious content be discovered?