The Fake Project is a research project of the Institute for Informatics and Telematics (IIT) of the Italian National Resarch Council (CNR), which aims to study the phenomenon of Fake Twitter Followers in a scientific manner.
Building on a baseline dataset of Twitter accounts (in which fake and real accounts are known a priori) and on a series of criteria, detailed in a conference paper and in a technical report, we developed a fake detection engine.
Some of the preliminary results we have obtained with our engine are reported below.
Twitter Account | Followers Count | Fake Percentage | Inactive Percentage | Genuine Percentage | |
---|---|---|---|---|---|
Barack Obama | @BarackObama | 41 Millions | 8.5% | 57.1% | 34.4% |
François Hollande | @fhollande | 608 Thousands | 5.3% | 63.6% | 31.1% |
David Cameron | @David_Cameron | 595 Thousands | 11.7% | 24.0% | 64.3% |
Should you want to try the engine, we will be glad to give you the credentials to access the detector!
Please, send us an email to fakeproject@iit.cnr.it, and we will soon reply. In the meantime, you can have a look to the classifier demo here.
The research leading to the Fake Follower detection engine has been partially supported by the Registro.it funded projects My Information Bubble and SoS - Social Sensing, and by the regional project MyChoice.
Many thanks for your interest in the project,
The Fake Follower Classifier Team