Abstract

In this research, an enforceable security policy model which can prevent the browser from visiting malicious URL was designed, simulated, and evaluated. This is with a view to enhancing security performance of browsers. The main aim of this study is to develop enforceable security policies for browsers and specific objectives are to design a malicious detection model, implement the model and evaluate the performance of the model. The model was designed using the Petri-net model and also, in the study, browser hound (Bhound) was designed and simulated using the stringmatching algorithm which serves as a control measure to check for URLs similar to the filters stored in the local storage. String matching algorithm was employed for the compromised URL detection model and the model was evaluated using performance metrics; such as accuracy, precision and recall rate. The results of system simulation were found to have a 91.2% overall detection rate, 90.0% precision, 28.8% false-positive rate, and 86.0% classification rate. The software development model adopted was a scrum, and the implementation was done using JSON, JavaScript, and hypertext mark-up language (HTML). This system developed is capable of enhancing browser security and could as well handle typosquatting which is one of the schemes used by attackers.