AnonShield: Scalable On-Premise Pseudonymization for CSIRT Vulnerability Data
Researchers have developed AnonShield, a system designed for scalable, on-premise pseudonymization of vulnerability data used by computer emergency response teams (CSIRTs). This system utilizes GPU-accelerated named-entity recognition (NER), streaming processing, and caching to significantly reduce processing times, achieving up to a 738x speedup on datasets up to 550 MB. AnonShield demonstrates that effective pseudonymization is achievable without compromising the analytical utility of the data, facilitating compliant data sharing within operational CSIRT environments. AI