This study investigates (i) the web articles related to DDoS attacks to summarize their causes, (ii) quantifies subsequent losses, and (iii) proposes mitigation strategies. We propose a text mining-based Cyber-risk Assessment and Mitigation (TCRAM) model comprising three modules. Firstly, the cyber-risk assessment (CRA) module analyzes textual web articles and extracts themes using Latent Dirichlet Allocation. Subsequently, we estimate the probability of misinterpreting these themes using the kernel Naïve Bayes classifier. Next, the cyber-risk quantification (CRQ) module calculates the expected loss incurred by a firm due to DDoS attacks. Lastly, the cyber-risk mitigation (CRM) module aids the CTO in reducing, accepting, or passing the cyber-risk. Our CRA module observes that hackers use obsolete protocols to launch DDoS attacks. While, the CRQ module highlights that losses are dependent on attack size and duration based on risk theory. We also note that our model mostly misclassifies attack features and cost topics. The CRM module suggests using analytics-based mitigation strategies to reduce the cyber-risk or pass it to cyber-insurers. Based on risk theory, our framework helps CTOs invest appropriately in technology and cyber-insurance.

Abstract Only