Nowadays, Data Sanitization is considered as a highly demanded area for solving the issue of privacy preservation in Data mining. Data Sanitization, means that the sensitive rules given by the users with the specific modifications and then releases the modified database so that, the unauthorized users cannot access the sensitive rules. Promisingly, the confidentiality of data is ensured against the data mining methods. The ultimate goal of this paper is to build an effective sanitization algorithm for hiding the sensitive rules given by users/experts. Meanwhile, this paper concentrates on minimizing the four sanitization research challenges namely, rate of hiding failure, rate of Information loss, rate of false rule generation and degree of modification. Moreover, this paper proposes a heuristic optimization algorithm named Self-Adaptive Firefly (SAFF) algorithm to generate the small length key for data sanitization and also to adopt lossless data sanitization and restoration. The generated optimized key is used for both data sanitation as well as the data restoration process. The proposed SAFF-based algorithm is compared and examined against the other existing sanitizing algorithms like Fire Fly (FF), Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Differential Evolution algorithm (DE) algorithms and the results have shown the excellent performance of proposed algorithm. The proposed algorithm is implemented in JAVA. The data set used are Chess, Retail, T10, and T40.

中文翻译：

---

用于改进数据清理和恢复的自适应优化

如今，数据净化被认为是解决数据挖掘中隐私保护问题的一个高度需求的领域。数据清理，是指用户给出的敏感规则进行了特定的修改，然后释放修改后的数据库，使未经授权的用户无法访问敏感规则。很有希望的是，数据的机密性是根据数据挖掘方法来确保的。本文的最终目标是建立一种有效的清理算法来隐藏用户/专家给出的敏感规则。同时，本文专注于最小化四个清理研究挑战，即隐藏失败率、信息丢失率、错误规则生成率和修改程度。而且，本文提出了一种启发式优化算法，称为自适应萤火虫（SAFF）算法，用于生成小长度密钥进行数据清理，并采用无损数据清理和恢复。生成的优化密钥用于数据清理和数据恢复过程。将所提出的基于 SAFF 的算法与其他现有的消毒算法如 Fire Fly (FF)、遗传算法 (GA)、粒子群优化 (PSO) 和差分进化算法 (DE) 算法进行了比较和检查，结果显示了出色的提出的算法的性能。所提出的算法在 JAVA 中实现。使用的数据集是 Chess、Retail、T10 和 T40。