Graphing Website Relationships for Risk Prediction: Identifying Derived Threats to Users Based on Known Indicators

The Graphing Website Relationships for Risk Prediction whitepaper researches the relationship based on referrer links and the number of hops to a malicious site. Cybersecurity products provide indicators to threats, but actions on risk indicators are executed singularly without evaluating the relationships. Enhancing the performance of existing indicators with threat levels could provide users with additional information about the potential risk during online browsing.

The purpose of the research was to test a method of identifying risks to websites based on the relationship to known web browsing security threats. Since a lag exists between identifying an instance of malware and propagation of the threat notifications, blacklists cannot provide a complete solution to identifying risks. The rationale for the study was to test whether a risk could be predicted for non-malicious websites based on the proximity to malicious sites as determined by known threat indicators. In the current research, we sought a method of enhancing the existing threat indicators with additional metadata about the evaluation of risk based on relationships. By delineating threats based on the relationships to other websites, organizations could use the information to determine the threshold for acceptable risk. The acceptable risk would be based on the probability that a user visiting a website may follow a path to malicious content.

Download the full whitepaper to view the results, methodology, and data for the Graphing Website Relationships for Risk Prediction study.