Trust Evaluation of Doctor Behavior Based on GCN Network in Medical Big-Data Access Control

Abstract

The establishment of health care big data has brought great convenience to population health and medical research, but at the same time a series of privacy protection issues must be considered as a result. In this paper, we propose a graphical convolutional neural network to detect the access behavior of doctors in medical big data. In this paper, we propose a graphical convolutional neural network to model the access behavior of doctors in medical big data and perform trust evaluation, so as restrict such doctors or behaviors. In this paper, by taking the doctor behavior features and the doctor-doctor relationship network as input, the GCN network is used to supervise the learning of the department to which the doctor belongs, and the last layer is used as the characterization learning result. Finally, the similarity between doctor and department is used as the doctor behavior trust evaluation index. The experimental results show that the proposed model in this paper can well identify doctors' behaviors with malicious intent and assign a low trust value, laying the foundation for further research.