A fourfold-objective-based cloud privacy preservation model with proposed association rule hiding and deep learning...

Area of Science:

• Computer Science
• Information Security
• Data Privacy

Background:

• Existing cloud privacy solutions struggle with sensitive data protection.
• A robust model is needed to ensure data confidentiality in cloud environments.

Purpose of the Study:

• To propose a novel privacy preservation model for cloud environments.
• To enhance the security of sensitive data against unauthorized access.

Main Methods:

• Sensitive data identification using Augmented Dynamic Itemset Counting (ADIC) based Associative Rule Mining.
• Generation of an optimal tuned key using a deep learning approach with LSTM and a hybrid optimization (MUAOA).
• Data sanitization using the generated tuned key and data restoration.

Main Results:

• The proposed LSTM+MUAOA model demonstrated superior privacy preservation.
• Achieved higher privacy values (5.21) compared to existing models in comparative evaluations.

Conclusions:

• The developed privacy preservation model effectively secures sensitive data in the cloud.
• The integration of ADIC, LSTM, and MUAOA offers a promising approach for advanced data security.