A Scalable Data Partitioning Framework to Enhance Cloud Security
Aditya Lal, Georgia Technology · International Journal of Engineering Research and · 2019
The huge data on the websites due to rapid use of cloud computing increase the security concern.Cloud computing provides various services such as storage, scalable transactions, sharing and downloading.There are many techniques which provides security of these services but these are dynamic in nature and leads to overload the memory.In this paper, a novel framework has been presented based on data partitioning technique which avoids the data duplication and provides scalable consistent data with enhanced security.The proposed work divided into three parts in which Dice Coefficient and Cosine Similarity has been used to secure the similar data.The data has been further encrypted using the encryption system to enhance the data integrity and security level.The encrypted data further consider as input for the data partition techniques in which different algorithms has been developed to partition the data in blocks.The proposed work further optimized using the Artificial Bee Colony algorithm to optimize the partitioned data.The obtained results further cross validated using the classification technique Artificial Neural Network (ANN) in which data has been trained and tested to check the effectiveness of the proposed work.The experimental results finally validated by measuring the True Positive Value (TP), False Negative Value (FN), True Negative Value, and False Positive value.The good TP and bad FN shows that proposed framework efficiently partitioned the data in an effective and scalable way.