This article aims to systematize approaches to selecting predictive models based on associative search and analyze their effectiveness and applicability in the operating conditions of nuclear power plants (NPP). The study examines the main approaches and algorithms used for associative data analysis. In addition, the article includes a review of existing work in this area, evaluates its results, and suggests new promising directions for further research. The article is devoted to the issue of selecting a predictive model based on associative search for a technological process (TP) parameter to predict the risk potential of the TP of an NPP power unit using the example of one of the parameters of the information task of the upper level control system of the automated process control system of an NPP - cooling water heating.