Predicting Cyber Events by Leveraging Hacker Sentiment

Ashok Deb,Kristina Lerman,Emilio Ferrara

DOI: https://doi.org/10.3390/info9110280

2018-04-15

Abstract:Recent high-profile cyber attacks exemplify why organizations need better cyber defenses. Cyber threats are hard to accurately predict because attackers usually try to mask their traces. However, they often discuss exploits and techniques on hacking forums. The community behavior of the hackers may provide insights into groups' collective malicious activity. We propose a novel approach to predict cyber events using sentiment analysis. We test our approach using cyber attack data from 2 major business organizations. We consider 3 types of events: malicious software installation, malicious destination visits, and malicious emails that surpassed the target organizations' defenses. We construct predictive signals by applying sentiment analysis on hacker forum posts to better understand hacker behavior. We analyze over 400K posts generated between January 2016 and January 2018 on over 100 hacking forums both on surface and Dark Web. We find that some forums have significantly more predictive power than others. Sentiment-based models that leverage specific forums can outperform state-of-the-art deep learning and time-series models on forecasting cyber attacks weeks ahead of the events.

Computation and Language