Scalable PII Discovery in Mobile App Databases via Hypothesis-Driven Search
A hypothesis-driven framework that treats PII localization as bounded, adaptive search under uncertainty, and separates lightweight PII exploration from targeted extraction, normalization, and deduplication over validated regions, thereby limiting exhaustive inspection to regions supported by sampled evidence.