东盟数据匿名化指南（英）-2025.1_53页_2mb

Anonymization is a risk-based process for converting personal data into non-identifiable data, providing benefits like compliance, trust, and safe data sharing. It involves removing or altering data to prevent re-identification. Key terms include personal data (identifiable data), anonymized data (low re-identification risk), direct identifiers (high risk), and indirect identifiers (medium risk). The process includes five steps: know your data (assess suitability and de-identify); apply techniques (like suppression, masking or pseudonymisation); compute risks (using thresholds like k-anonymity); and manage risks (with controls). k-anonymity measures resistance to linking attacks, while l-diversity and t-closeness address additional risks for target attributes. Common misunderstandings highlight insufficient obfuscation, and tools like PDPC's Excel-based tool assist in implementation. Benefits include enhanced privacy and data utility, but risks depend on use cases and data sensitivity.