A Dynamically Configurable Document Protection Framework for Data Leakage Prevention Using Extended Host-Based Scrutiny and Hybrid Fingerprinting
Keywords:
Data leakage prevention, insider threat, document protection, EHDS, fingerprinting, access control, behavioural monitoringAbstract
This paper examines a dynamically configurable document framework that integrates policy-based access control, cryptographic protection, content fingerprinting, and host-based behavioural monitoring. The design incorporates a Probabilistic Incremental Algorithm (PIA), an Autoregressive Hidden Markov Model (AHMM), and an Extended Host-Based Data Scrutinizer (EHDS). Previously reported transaction experiments indicate true-positive rates of 98.1–98.3% for EHDS and 82.1–92.2% for alignment and sampling, with false-positive rates of 9.5–11.0% and 41–45%, respectively. Reported component execution times are also lower for EHDS. Two allocation scenarios illustrate selective distribution under configurable policies. These findings motivate further investigation of integrated document protection; they do not constitute an independent experimental validation. Reproducible algorithm specifications and real-world evaluation remain necessary.