An Adaptive Access Control Model Based on Trust and Risk for Medical Big Data

Although big data technology is widely used in the medical field, there is a risk of privacy leakage in the process of medical data share. In terms of this problem, firstly, this paper divides user access behavior into two stages: selecting work objects and selecting what kind of which information to visit; then uses information entropy to measure the confusion degree when users select objects in the process of access; and uses entropy as the input of BP neural network to quantify the risk of access behavior. The risk is fed back to the trust center, which calculates the value of user trust according to the principle that the value of user trust is negatively correlated with the risk. If the user trust value does not meet the requirements of the assigned role, the role will not be assigned. At the same time, if the risk quota is exhausted, the access will be blocked. In this way, access control of medical data is realized.