A Comprehensive Dataset for Context-Aware Security Monitoring for Anomaly Detection

This article describes a high-dimensional dataset designed for context-aware security monitoring and anomaly detection in digital transaction systems. The data encompasses 112 features, including transactional metadata, user behavioral patterns, and advanced technical telemetry (IP reputation, VPN detection, device fingerprinting, network security indicators). A unique aspect of this dataset is the integration of Large Language Model (LLM) outputs, providing risk scores and natural language reasoning for contextual security assessment. This dataset is particularly valuable for training machine learning models that require a hybrid approach, combining traditional tabular data with AI-generated contextual insights to identify security threats such as account takeover, suspicious behavioral patterns, and social engineering attacks across multiple digital channels.