Cybersecurity: Network models of SPAM detection by humans

Post date: Nov 1, 2018 9:41:49 PM

We are developing and testing Bayesian Network models that describe the psychological representation of SPAM and phishing e-mails. Understanding how people classify and make decisions about e-mails based on their observable features is an important step toward addressing security problems due to mistakes made by e-mail and Internet users.

We are exploring a principled "object oriented" approach to theory development, as well as building "influence diagrams" that model the decision process, including utility judgments which appear to moderate willingness to engage in risky behaviors.