Formula omitted]: Keyword-based auditing with frequency hiding and retrieval reliability for smart government

Abstract

In a keyword-based auditing paradigm, users typically focus on specific parts of a dataset rather than the integrity of the dataset as a whole. However, this approach is subjective and can be exploited by malicious storage servers to analyze audit frequency or reduce backups. This poses a significant risk for government data, where any privacy leakage or corruption could have catastrophic consequences. In this paper, we propose a keyword-based auditing scheme for smart government called KA, which provides both frequency hiding and retrieval reliability. KA leverages a Bloom filter to adjust the false positive rate and audits files based on specified keywords and random files obtained through fuzzy matching. To obtain privacy-preserving fuzzy matching, KA constructs an index table embedded with update times to retrieve a wide range of files to be audited. This approach is secure against replay attacks and supports index table updates through structure iteration instead of recalculation. Additionally, KA uses a relation matrix to detect all challenged storage nodes and ensure honest storage proof. KA provides storage robustness, privacy protection of hidden frequencies, Data security and retrieval reliability. Furthermore, It reduces audit computation overhead by 32.6% compared to probabilistic public auditing.