The paper describes the approaches used in the usability assessment of user interfaces based on the idea of searching patterns in user behavior. The existing research in this domain is described. It is shown that the task of recognizing patterns of user behavior can be represented as a task of searching for sequential patterns of Data Mining. We propose a user activity model for the recognition of event patterns. By event, we mean a fact recorded at a certain time during the interaction of a certain user on a certain device with the software user interface. This model can be used in the analysis of the usability of user interfaces and can be used to solve problems of increasing the efficiency of user interaction with information systems. Also, we propose an algorithm for the recognition of repetitive event patterns, which uses the methods of Data Mining (association rules and sequential patterns). Testing of this algorithm showed that it allows experts to detect previously unknown patterns of user interaction with information systems at a lower time cost.