Privacy-preserving Search over Encrypted Personal Health Record in Multi-Source Cloud

Abstract

The biggest domain which needs a cloud platform in this smart world is the medical domain, where hospitals have a huge amount of data regarding the patient’s records. When these personal health records are stored in the cloud, the system needs some encryption mechanism to secure the records. In this paper, an enhanced Attribute-Based Encryption (ABE). The health records of the patient are decrypted with the help of private keys. In this model, the patient doesn’t need any certificate for authorization, so the model uses the details of the patients as attributes, and these details are encrypted. A threshold value is defined as t, if the number of encrypted attributes is matched with the threshold value then only the data owner can decrypt the data. A hierarchical mechanism helps the model to have smooth access to the data. The traditional algorithms are good at handling a small amount of data but this algorithm provides security as well as scalable property. The enhanced mechanism includes the unsupervised machine learning approach to label the patients based on their disease. In the cloud, a group of clusters are formed to separate the patients and allocate the doctors based on their disease treatment. The process of labeling is major drawback in case of attribute based encryption mechanism.