Social Engineering Attacks Detection and Mitigation by Analyzing Human Interaction Patterns

Social engineering is used by cybercriminals as a deceitful tactic, to trick people into giving away private information. It is achieved by granting them unauthorized access, or doing activities that can jeopardize their personal security. So, this paper provides an insight in to the analysis of human interaction patterns to identify and counterfeit these kinds of attacks. Social engineering attacks are dynamically changing techniques, so analyzing various parameters like behavioral analysis, network parameters, access patterns, and underlying machine learning algorithms is crucial. This work also provides the comparison of various machine learning algorithms like adaptive boosting, support vector machine, logistic regression, voting, random forest, multilayer perceptron, Naïve Bayes, decision trees, and nearest centroid.