This article provides a review of publications on the analysis of students’ satisfaction with the educational process based on natural language processing methods. 197 student feedback on 129 elective disciplines at University of Tyumen was collected. A comparative analysis of keyword extraction methods was conducted: statistical TF-IDF, RAKE and YAKE; contextual KeyBERT; graph-based TextRank. On the collected reviews, grouped by elective disciplines, the RAKE method had the highest F1 BERTScore with 79 %. By parsing open sources, a dataset with 2210 Russian-language reviews for courses of different educational platforms was formed. Machine learning methods for sentiment analysis were described: support vector machines, logistic regression and based on Transformers, comparison on the manually marked part of the collected reviews. After fine-tuning on the rubert-base-cased model macro-averaged F1- score became 71.6 %. Classification into three classes (negative, neutral, positive) is not performed for the whole text of the review, but separately for each sentence from that text. The implementation of a database and information system for collecting and analyzing student feedback on the studied elective courses are presented. The model for sentiment analysis of the feedback is put into a separate microservice, which is communicated through an interface of the freely distributed Python framework FastAPI. The information system is designed to help students choose electives based on more qualitative data, and teachers and university administration ‑ to draw conclusions for further transformation of the educational space, taking into account students’ opinions.

The process of capturing wave energy by an island around which the bottom has a conical shape is studied. The kinematics of the initially straight wave front near such an island was investigated by the method of step-by-step orthogonal advancement. An estimate of the part of wave energy that due to refraction is captured by the bottom slope surrounding the island was obtained. Numerical modeling of this problem was also carried out within the framework of the shallow-water model, which confirmed the results obtained by the kinematic method, including a quantitative assessment of the part of the wave energy reflected by the island and captured by its inclined shelf. It is shown that the islands surrounded by the bottom slope well shield the water area located behind such an island.

Year by year researchers use the method of electrotomography more extensively to solve a wide variety of tasks. For example, electrotomography can be applied in archaeological excavations, in the tasks of controlling mine tailings, in engineering surveys, to study fault structures, for monitoring studies in seismically active areas. It is necessary to perform sufficiently long-term observations to form approaches in solving the problem of predicting seismic events. This leads to the need to consider large arrays of initial data and interpret a significant amount of field data. In this regard, it is important to use and develop modern computer tools for processing and interpreting the results of regular observations. The purpose of this work is to modernize and develop the Direct-Inverse-Solver (DiInSo) software package for solving direct and inverse problems for processing, interpreting and analyzing electrotomography monitoring data.

Nowadays, various technologies are used to create intelligent assistants, based both on research in the field of neural networks, natural language text analysis tools, and on the use of semantic modeling tools. Each of these approaches allows you to qualitatively solve certain problems. As part of this work, an intelligent assistant is being developed that combines all these approaches. The purpose of the work is to create an intelligent assistant that performs as a virtual consultant on the organization’s work processes. The development is based on the use of the semantic model of the organization and business processes. To recognize user intents, we use homomorphic and generalized user intents. The system allows decomposing user tasks and creating a consistency of their execution based on the user semantic models and the subject area.

Analysis of modern approaches to the implementation of driver assistance systems, as well as the implementation of the architecture of the driver assistance system, aimed at recognizing traffic signs at the maximum distance from it under difficult weather conditions, for early feedback to the driver. The paper considers the main signals used in the implementation and operation of the driver assistance system: data from the car's CAN bus, information from a GPS receiver, video fragments from a digital camera. The presented modular architecture uses the listed data sources for estimating the traffic situation, as well as neural network methods for recognizing traffic signs. The modular architecture of the driver assistance system is presented, which allows notifying the driver about traffic signs. The system is equipped with lane boundary control to alert the driver to signs related to the adjacent carriageway when turning. It has been experimentally proven that the modular architecture of the driver assistance system presented in the paper is not inferior in speed and accuracy to alternative systems, acting as a comprehensive autonomous solution.