KIMONO_FakeNews_shutterstock_1481950463.png

DEU-ISR Research Cooperation: KIMONO

Campaign Identification, Monitoring and Classification Using Social Media Mining Methods for Integration in an AI-based Early Warning System

(2021 – 2024)

Influence and disinformation campaigns can manipulate and influence people and thus lead to a massive weakening of trust in democracy and its constitutional principles as well as trust in freedom of expression. In addition, there is a risk that state actors will use influencing campaigns in social media specifically to destabilize opposing states (hybrid warfare). It is therefore essential to recognize these campaigns as early as possible and to track their development and spread in order to ultimately be able to initiate suitable countermeasures. The focus of the research work is therefore on campaign identification, monitoring and classification using social media mining methods for integration into an AI-based early warning system.

Tasks and objectives

  • Software support to not only recognize and explicitly flag short-term campaigns, but to provide additional explanations as to why a particular social media post is classified as part of a misleading information and influence campaign.

  • Support for social media analysis with a focus on recognizing opinions. In addition, the software should be adapted to specific tasks and enable integrated visualization based on a graph database.

  • Provision of a catalog of requirements for the subsequent development of an early warning system.

 

With funding from the:

logo-federal-ministry-of-defence.jpg