IIT Home Page CNR Home Page

Fame for sale: effcient detection of fake Twitter followers

Fake followers are those Twitter accounts speci cally created to inflate the number of followers of a target account. Fake followers are dangerous for the social platform and beyond, since they may alter concepts like popularity and in influence in the Twittersphere|hence impacting on economy, politics, and society. In this paper, we contribute along diff erent dimensions. First, we review some of the most relevant existing features and rules (proposed by Academia and Media) for anomalous Twitter accounts detection. Second, we create a baseline dataset of veri ed human and fake follower accounts. Such baseline dataset is publicly available to the scientifi c community. Then, we exploit the baseline dataset to train a set of machine-learning
classi ers built over the reviewed rules and features. Our results show that most of the rules proposed by Media provide unsatisfactory performance in revealing fake followers, while features proposed in the past by Academia for spam detection provide good results. Building on the most promising features, we revise the classi ers both in terms of reduction of over fitting and cost for gathering the data needed to compute the features. The final result is a novel Class A classi er, general enough to thwart overfi tting, lightweight thanks to the usage of the less costly features, and still able to correctly classify more than 95% of the accounts of the original training set.
The findings reported in this paper, other than being supported by a thorough experimental methodology and interesting on their own, also pave the way for further investigation on the novel issue of fake Twitter followers.


2015

Autori IIT:

Angelo Spognardi

Foto di Angelo Spognardi

Tipo: TR Rapporti tecnici
Area di disciplina: Information Technology and Communication Systems
IIT TR-05/2015

File: TR-05-2015.pdf

Attività: Social Media Analysis