Towards an architecture for monitoring communications in social networks based on graphs -using honeypot
Kanga Koffi, Kamagaté Beman Hamidja, Pacôme Brou, Olivier Asseu, Souleymane Oumtanaga · International Journal of Basic and Applied Sciences · 2024
In this paper, we are proposing an architecture for monitoring communications in social networks. The main objective is to establish a sys-tem that can detect possible attempts to intrude communications in a social network environment using honeypots. To do this, we reviewed the various works related to intrusion detection concerning architectures, algorithms and software tools in this area. Concretely, our proposal includes four (4) components that we have presented while defining the role of each of them. We implemented this architecture in a python environment with the associated algorithms: One-Class SVM - FOREST ISOLATION as well as their combination. The results show that our architecture produces a more refined level of intrusion detection by applying combinations of these different algorithms. Anything to ensure that intrusions detected by honeypots would be reliable using our proposal.