IT Support Tickets Dataset for Social Engineering Detection

A dataset of 40,000 IT helpdesk support tickets, each labelled as either malicious (social engineering attack) or legitimate. Each ticket includes the raw text content and engineered linguistic features — urgency markers, authority references, first-person pronoun counts, question marks, and average word length — alongside metadata such as category, complexity, urgency level, and timestamp. The dataset is designed for binary classification tasks aimed at detecting social engineering attempts embedded in routine IT support requests.