An Expert System for the Detection and Mitigation of Social Engineering Attacks (Sea) Using Machine Learning Algorithm

Abstract

Social engineering is a mechanism of convincing an Internet user to disclose their private information. It is a technique of influencing people, so that they pretend to share their private information. Using social engineering attack (SEA) individual’s or the systems are targeted by the criminals, trying to access the password and bank information’s or secretly install malicious software, this also gain access to your computer. The attackers not only capture an individual’s information, also access information in various business environments such as discussion with the corporate clients, interaction between the employees through a chat service. This can be identified, and its effect can be analyzed using advanced machine learning models. In the research work we have proposed and demonstrated the implementation of a model using machine learning, to analyze the malicious behavior of the attackers and for the detection of SEA. This method is applied for both off-line text and in real-time situations to identify whether a human Chatbot or off-line chat is performing a SEA or not. In the proposed method the text message is analyzed and verified for linguistic errors and the model used to detect and separate the chances of any attacks. The technique proposed is also analyzed using both factual and semi-synthetic dataset to obtain the result with better accuracy. The model developed will detect phishing links (uniform resource locator -URLs) from the webpage and classify the URL whether it is legitimate or illegitimate URL using machine learning algorithm.