About

This project, conducted as a collaborative effort by myself and three fellow engineering students (Nathan DUNAND, Clément LAISNÉ, Corentin HUVELIN), represents our fourth-year engineering school project in the field of cyberdefense.

This article presents an engineering project aimed at automating the detection of information manipulation, specifically focusing on the field of computational influence operations. These operations involve actions taken to manipulate public opinion through digital media and social networks. The objective is to develop a tool capable of detecting online disinformation using a dataset.

The project relies on indicators from scientific literature, including word count, tweet length, hashtag frequency, sentiment score, and others. A three-step methodology involving data processing, visualization generation, and comparative analysis aids in identifying suspicious accounts and datasets. Additionally, potential enhancements such as expanding data collection across platforms and languages, and incorporating machine learning for account classification, are discussed.

For a detailed analysis and further insights, please refer to the original article.