BDFIS: Binary decision access control model based on fuzzy inference systems

Abstract

Access control is a ubiquitous feature in almost all computer systems, and as data becomes more and more of an important asset for organizations, so do the associated access control policies. However, with the increase in the amount of data being produced, e.g. in IoT and social networks, the interest in simpler access control is increasing as well since more subjects (public, researchers, etc.) are now requesting access to it. Defining the exact conditions to allow each subject to access the data can be difficult, especially when vaguely defined conditions such as "expertise of a researcher" come into play. Fuzzy Inference Systems (FIS) allow to process these vague conditions and enables access control mechanisms to be more easily applied. The contribution of this paper lies in showing how a FIS can be used to output binary access control decisions (grant/deny) and what are the differences in the inference process that stems from restricting the output to these two output values.