Kernel naive Bayes classifier-based cyber-risk assessment and mitigation framework for online gaming platforms

Recently, the number and intensity of cyberattacks against massively multiplayer online (MMO) gaming platforms have increased; up to 74% of distributed denial-of-service (DDoS) attacks on MMO gaming (MMOG) firms have been launched by hackers. These malicious attacks affect gamers' experience and MMOG firms' revenue model. Along with financial losses, MMOG firms' reputation also suffers from these attacks. Therefore, in this study, we devised a framework to quantify and mitigate cyber-risk for MMOG firms using a hybrid learning method, namely, a kernel naive Bayes classifier. Our kernel naive Bayes classifier-based cyber-risk assessment and mitigation (KB-CRAM) framework included the DDoS attack traits. Subsequently, it outputs (i) the probability of DDoS attacks; (ii) the expected financial losses; and (iii) cyber-risk mitigation strategies, such as self-protection (technology, compliance, and legal deterrence), self-insurance, or cyber-insurance. Our study contributes to field-relevant literature by providing managers with a tool to improve game performance. This framework also suggests ways in which MMOG firms can hedge losses against repeated attacks from unethical hackers.

Automated summary

This study introduces a framework for quantifying and mitigating cyber-risk for massively multiplayer online gaming platforms using a kernel naive Bayes classifier. The framework includes DDoS attack traits and strategies for risk mitigation, providing managers with tools to improve game performance and hedge against repeated attacks.