Open source intelligence gathering and analysis of cyber attack trends

Abstract

With the emergence and growing dominance of the Internet, the cyber threat landscape has experienced rapid changes in recent years. As people struggle to understand and keep up with the latest threats, the lack of readily available resources is a challenge faced by many. To address this issue, proper intelligence gathering must be done, where subsequent analysis work can allow us to understand the changes in the world better.

In the project, publicly available repositories are compiled using different open-source intelligence (OSINT) techniques, from the repository we were able to identify that the healthcare industry are more susceptible to cyber incidents. By using different machine learning models such as K-Nearest Neighbour and MLPClassifier, prediction of economic impacts and prediction of attack type is done. We find that with a structured repository available, we were able to predict the attack type of a cyber incident to an accuracy of 41.53%, and the KNN model used for the prediction of the economic impact attains the best results when k-value = 14.