Застосування інтелектуального аналізу текстів для вирішення задач інформаційної безпеки. Частина 2. Пошук інсайдерів

Abstract

An algorithm for detecting insiders based on automated semantic analysis of sent emails is proposed. In the Orange software environment, based on the database "The Enron Email Dataset", a study of the possibilities of classifying the organization's personnel into insiders and ordinary employees was carried out. The accuracy of identifying insiders was 60%