This paper proposes a retrievable data perturbation model for overcoming the challenges in cloud computing. Initially, genetic whale optimization algorithm (genetic WOA) is developed by integrating genetic algorithm (GA) and WOA for generating the optimized secret key. Then, the input data and the optimized secret key are given to the Tracy–Singh product-based model for transforming the original database into perturbed database. Finally, the perturbed database can be retrieved by the client, if and only if the client knows the secret key. The performance of the proposed model is analyzed using three databases, namely, chess, T10I4D100K and retail databases from the FIMI data set based on the performance metrics, privacy and utility. Also, the proposed model is compared with the existing methods, such as Retrievable General Additive Data Perturbation, GA and WOA, for the key values 128 and 256. For the key value 128, the proposed model has the better privacy and utility of 0.18 and 0.83 while using the chess database. For the key value 256, the proposed model has the better privacy and utility of 0.18 and 0.85, using retail database. From the analysis, it can be shown that the proposed model has better privacy and utility values than the existing models.

中文翻译：

---

Tracy–Singh产品和遗传鲸鱼优化算法，用于可检索数据扰动，用于云计算中隐私保留的数据发布

本文提出了一种可克服的数据扰动模型，以克服云计算中的挑战。最初，通过集成遗传算法（GA）和WOA来生成优化的密钥，从而开发了遗传鲸鱼优化算法（genetic WOA）。然后，将输入数据和优化的密钥提供给基于Tracy-Singh产品的模型，以将原始数据库转换为受干扰的数据库。最后，当且仅当客户端知道秘密密钥时，客户端才能检索受干扰的数据库。基于性能指标，隐私和效用，使用FIMI数据集的国际象棋，T10I4D100K和零售数据库这三个数据库来分析所提出模型的性能。此外，将提出的模型与现有方法进行了比较，例如对于键值128和256的可检索通用附加数据扰动，GA和WOA。对于键值128，在使用国际象棋数据库时，所提出的模型具有较好的隐私性和实用性，分别为0.18和0.83。对于键值256，使用零售数据库，建议的模型具有更好的隐私性和实用性，分别为0.18和0.85。从分析中可以看出，所提出的模型比现有模型具有更好的隐私和实用价值。