Bit Incorporated Codon Positional Encoder (BICPE) Algorithm for Privacy Preservation in Vertically Partitioned Data in Cloud

HIGHLIGHTS

• Fake transactions are inserted into database to hide item frequency.

• The input binary bits are grouped into N sets with length as four for encryption.

• Each item in a data set is encoded with A,C,G,T DNA codons.

• Encrypted Association rules are built for frequent items in a transaction.

Abstract

Current days an enormous amount of data is put away and recovered from the cloud because of its benefits like stowage size, assets pooling, and multi-occupancy. Because every administration is accessible through cyberspace, there is a lot of ambiguity about security and protection at various levels. This work proposed a privacy-preserving technique by applying Deoxyribonucleic Acid (DNA) coding-based novel Bit Incorporated Codon Positional Encoder (BICPE) algorithm for vertically divided cloud information. The transactional items are scrambled with a replacement of the DNA codon. Fictitious exchanges are appended unilaterally into the private database before encryption by the data owner to alleviate frequency analysis attacks of the items. The proposed BICPE algorithm encrypts the transactional data into A, C, G, T codons to upgrade the safety of the transactional items by replacement, operation, and intricacy. The concepts of binary coding and random values are utilized in DNA sequencing to encode the data. In this method, both the sender and the receiver will share a common DNA sequence for encryption and decryption process. The performance of the proposed technique is compared with the traditional Homomorphic Encryption (HE) in terms of support threshold values (Ts) for different items in a transaction. The exploratory outcomes demonstrate that the BICPE algorithm outperforms HE as far as ciphertext size, execution time, and privacy. Hence, the newly proposed technique is more efficient and offers better performance in terms of privacy.

Keywords:
Frequent Itemset; Encryption; Eclat; Apriori; FP-growth; Association Rue Mining; DNA cryptography.