It is important to note, that within the last ten years we can see considerable growth in the number of incidents and revealed vulnerabilities as well as purposeful attacks on industrial automation and control systems with the purpose of industrial espionage, fraud, and violation of the company's functioning. In this paper, we offer a methodology that ensures the systematization of the main steps of the development of a profile and algorithms of adaptive authentication on the critical information infrastructure facility. The offered algorithm eliminates the disadvantages of existing traditional methods of authentication based on the use of explicit verification methods connected to the fact that the authentication characteristics which can be compromised by malefactors are applied to the determination of the user's identity. In the given article, we examine an example of implementation of the method with the application of a data qualifier on the example of a test platform. For classification, a machine learning model - the k-nearest neighbor algorithm - is used. For confirmation of the overall performance of the algorithm, an assessment of the robustness of the model according to the Euclidean metric and Chebyshev distance was used as well as an assessment of the descriptiveness of entry tags influencing the classification.