Retraction Notice: A Metric for Measuring Distances Between Attack Trees in Cybersecurity

In this paper, we explore a novel problem in cybersecurity: determining the distance between two attack trees. Attack trees, as initially described by Bruce Schneier, are rooted acyclic structures that model the various ways an attack can occur. We define attack trees using a formal recursive structure and propose a method to measure the distance between them. Our primary research goal is to establish a mechanism for calculating this distance, which can enhance threat modeling by allowing practitioners to compare and share attack trees more effectively. We outline the theoretical and usability requirements for our distance measure, ensuring it meets the criteria of being a metric, accommodating noisy human-developed labels, and incorporating all aspects of the trees. The common structure for sharing attack trees is defined by the ADTool XML Schema, while the means of comparison is developed through our proposed distance measure. This method can potentially identify similar threats, highlight missing components in analyses, and facilitate comparisons between human-drawn and machine-generated attack trees. Our approach aims to make threat modeling more robust and collaborative, providing practitioners with a valuable tool for threat analysis.