An Electronic Medical Record Access Control Model Based on Intuitionistic Fuzzy Trust
Rong Jiang,Rui Liu,Tao Zhang,Weiping Ding,Shenghu Tian
DOI: https://doi.org/10.1016/j.ins.2023.120054
IF: 8.1
2023-01-01
Information Sciences
Abstract:Electronic Medical Records (EMR), as the core data of medical big data, has improved the efficiency of medical services in the process of sharing and use, but a series of data privacy leakage problems have emerged. Access control technology can ensure users' legitimate and appropriate access to data resources, but traditional access control has many limitations such as static and coarse-grained, which is difficult to meet in the dynamic and complex healthcare environment. We create a model to solve the privacy leakage problem of healthcare big data. The model utilizes the Latent Delicacy Allocation (LDA) topic model to quickly search documents and extract the topic probability distributions between the target patient and other patients' electronic medical records, which improves the efficiency of medical record document analysis; Calculating physician access similarity, we quantify physician access behavior to more accurately describe physician access behavior in the form of data. We combined the intuitionistic fuzzy theory with the trust evaluation method to create a novel trust evaluation method-- Intuitionistic Fuzzy Trust Evaluation Method, which can reflect the uncertainty characteristics that exist in the physician's diagnostic and treatment process when quantifying the trust of the access by using this method, and it is more flexible and practical, which enhances the accuracy of the quantification of the trust. Improving the traditional trust aggregation method by adding a time window and time decay function, the optimized method improves the accuracy of trust aggregation, and the experimental results are more in line with the actual law. We created a new intuitionistic fuzzy trust access control model using a relative entropy-based intuitionistic fuzzy clustering algorithm and a reward and punishment mechanism to achieve adaptive and dynamic access control for physician access behavior. Experiments show that the access control model based on intuitionistic fuzzy trust proposed in this article achieves 95% correctness in identifying user access behaviors and 94.29% correctness in identifying excessive user access behaviors. Experiments have proved that this model has obvious advantages over other models in the process of privacy protection of electronic medical record systems and has certain control effects.