A Privacy Preserving Model for Enhancing the Public Access of Healthcare Data Using Hybrid Optimization and Data Anonymization Techniques

Abstract

The main demand of healthcare is sharing patient data with several parties [3]. Nevertheless, healthcare data contain sensitive information about individuals also publish such data is violate privacy. In this research work, design a Hybrid Strawberry and Golden Eagle based Recurrent Neural (HSGE-RN) with Quasi–Identifier Distribution Block (QIDB) anonymization scheme for enhancing public access by generating sensitive data. It involves attribute selection, fuzzy value generation, and QIDB Anonymization and the developed framework is implemented in Python tool. Initially, the healthcare unstructured dataset of various patients is collected and trained in the system. Then the collected dataset is updated to the designed design for converting the unstructured data into structured data using strawberry optimization. Hereafter, categorize the disease with fuzzy values based on the duration of a week, month, and year. After that, QIDB anonymization is employed for public access and secures the data of the patient. Finally, anonymized data are generated with the help of golden eagle optimization and the anonymized data are stored and shared with third parties. The attained outcomes of the designed model were validated and tested in terms of discernibility ratio, accuracy, classification error, running time, and scalability.