Two-Stage Neural Network Algorithm for Adapting Mobile Application Interfaces to User Features and Smartphone Characteristics

The article describes the algorithm for the formation of adaptive interfaces of mobile applications, based on the analysis of information about users and the characteristics of their smartphones. Describes the two-step approach to interface adaptation. At the first stage, the selection of each component of the prototype is implemented, taking into account the results of assessing the distinctive features of users and the technical characteristics of its mobile device. The second stage involves personalizing the interface taking into account user behavior. As a tool for implementing the developed adaptation algorithm, the use of a deep artificial neural network (INS) presented in the form of a multilayer perceptron is proposed. Input data, when operating INS, are the values of the formation of distinctive features of users (professional, psychophysiological, age, gender, as well as emotional state), output — components of the future interface. A software implementation of a neural network two-stage algorithm was developed when creating an adaptive interface of a mobile application of the training system. As a result of this application, a unique interface is generated for each student, generated taking into account its features. The introduction of the developed mobile application will provide a comfortable interaction between the student and the program, will reduce his emotional load, which will ultimately contribute to improving the efficiency and quality of training of university students.