Supervised Classification of Social Spammers using a Similarity-based Markov Random Field Approach

Abstract

Social spam has been plaguing online social networks for years. Being the sites where online users spend most of their time, the battle to capitalize and monetize users' attention is actively fought by both spammers and legitimate sites operators. Social spam detection systems have been proposed as early as 2010. They commonly exploit users' content and behavioral characteristics to build supervised classifiers. Yet spam is an evolving concept, and developed supervised classifiers often become obsolete with the spam community continuously trying to evade detection. In this paper, we use similarity between users to correct evasion-induced errors in the predictions of spam filters. Specifically, we link similar accounts based on their shared applications and build a Markov Random Field model on top of the resulting similarity graph. We use this graphical model in conjunction with traditional supervised classifiers and test the proposed model on a dataset that we recently collected from Twitter. Results show that the proposed model improves the accuracy of classical classifiers by increasing both the precision and the recall of state-of-the-art systems.

Detection of malicious social bots: A survey and a refined taxonomy

“…Overall, their approach showed 75% AUC score for early detection, increasing to above 95% after trending. -Mawass et al (2018) introduced a hybrid approach, in which graph-based method is coupled with machine learning classifiers in a probabilistic graphical model framework. They choose to use a graph based on the similarity between users' applications, rather than using the social structure of the network.…”

SimilCatch: Enhanced social spammers detection on Twitter using Markov Random Fields

“…This results in sybil accounts being linked, either through the social graph structure or through some form of similarity. This assumption is used to construct what can be called, depending on the application, a social ; , interaction Li et al (2016); Beutel et al (2013) or similarity graph Cao et al (2014); El-Mawass et al (2018). Detection is therefore executed either by means of graph clustering (or cutting) Cao et al (2014), or is modeled as a search for abnormally dense subgraphs Beutel et al (2013).…”

An Attention-Based Graph Neural Network for Spam Bot Detection in Social Networks

“…However, these methods cannot combine the features of tweet text. El-Mawass et al [17] linked similar accounts based on shared applications. They built an MRF model on the similarity graph, which used the similarity between users to spread information about their label belief.…”