A Technique for Securing Big Data Using K-Anonymization With a Hybrid Optimization Algorithm

Abstract

The recent techniques built on cloud computing for data processing is scalable and secure, which increasingly attracts the infrastructure to support big data applications. This paper proposes an effective anonymization based privacy preservation model using k-anonymization criteria and Grey wolf-Cat Swarm Optimization (GWCSO) for attaining privacy preservation in big data. The anonymization technique is processed by adapting k- anonymization criteria for duplicating k records from the original database. The proposed GWCSO is developed by integrating Grey Wolf Optimizer (GWO) and Cat Swarm Optimization (CSO) for constructing the k-anonymized database, which reveals only the essential details to the end users by hiding the confidential information. The experimental results of the proposed technique are compared with various existing techniques based on the performance metrics, such as Classification accuracy (CA) and Information loss (IL). The experimental results show that the proposed technique attains an improved CA value of 0.005 and IL value of 0.798, respectively.

An Extensive Study and Review of Privacy Preservation Models for the Multi-Institutional Data

“…Big data privacy has been successfully preserved using an algorithm known as Grey Wolf Optimizer-Cat Swarm Optimization (GWO-CSO). The generated model, which is used to create the k-anonymization database in which k numbers of duplicate records are created within the real database [17], is obtained by changing the update rules of GWO in the presence of the CSO algorithm. In order to provide safe data transmission over to end users, the newly created technique is used in the k-anonymized database to conceal the sensitive information relating to the data owners.…”

“…In order to provide safe data transmission over to end users, the newly created technique is used in the k-anonymized database to conceal the sensitive information relating to the data owners. The created technique ensures the k-anonymization phenomena by producing the parameters required to build the k-anonymized database ideally [17]. The privacy and utility metrics consider evaluating the fitness of the solutions obtained using the developed algorithm [17].…”

“…The created technique ensures the k-anonymization phenomena by producing the parameters required to build the k-anonymized database ideally [17]. The privacy and utility metrics consider evaluating the fitness of the solutions obtained using the developed algorithm [17].…”

“…Table 3, shows analysis concerning of journals and their affiliated papers. Journals from Communication cost [9] Search time in seconds [10] [23] Search time per identifiers in seconds [10] Communication overhead [13] [16] [20] Information loss [17] Classification accuracy [17] [21]…”

“…For example, Cisco expects that 500 billion devices will be connected to the internet by 2025 [14] [15] [16]. As a result, the amount of structured and unstructured data has increased by 2.5 Exa bytes daily [17]. On the other hand, the worldwide server farm's IP traffic would only arrive at 10.4 zetta bytes [18] [19] [20].…”