No abstract

Novoselova Ol'ga Viktorovna, associate professor of the department of general and applied physics

This article discusses methods for analyzing the information space using the "Sentiment Analysis" method. It is noted that an increased interest in tone analysis of information data of social network users began in the 2000s and continues to this day. Sentiment analysis has been studied in the works of both domestic and foreign scientists and studies. The article presents two popular methods for determining the sentiment of a text: unsupervised learning and the statistical method. Examples of each of them are given. An example of data parsing code is considered.Введение. В настоящее время социальные сети стремительно набирают обороты. На сегодняшний день уже более миллиона людей зарегистрированы в той или иной социальной сети, где они делятся информацией о своей жизни, интересах, об окружающем мире со своей аудиторией. Данные, которыми активно делится человек, могут быть использованы для определения интересов как бизнеса и государства, так и отдельной личности. Имея расширенные данные об интересах пользователя, можно определять различные отклонения в психике, определять потенциальных преступников, предотвращать различные нарушения со стороны пользователя социальной сети и прогнозировать конфликты.

The article discusses the issues of profiling users of social networks. Currently, according to statistics, more than half of the population of our country uses information technologies. Entertainment and interactive services such as social networks, blogs, forums, etc. have become particularly widespread among ordinary users. Such services are characterized by providing a lot of freedom to the user in entering various data about himself: the user can name himself as he likes, specify any place of residence, create a group with any description. The user creates the necessary content himself, searches for the necessary information himself, therefore, the search should be available to anyone, even someone unfamiliar with the skills of searching for data on a computer. Naturally, with the increase in the number of users, the availability of data decreases, it is necessary to implement more and more sophisticated search algorithms that allow people to find the information they need.As a rule, the search represents to the user a string in which he enters a query to which the program outputs the results found. The user is free to enter any data as a request, intentionally or accidentally distorting them, for example, by making typos. Users who create new content also often make inaccuracies in the names, distortions in words, add icons for decoration, write words in Latin. Thus, there may be errors in the data and in the queries, which must be taken into account when implementing the search.The article provides an overview of existing methods of fuzzy data retrieval. A review of existing studies has been conducted in which this problem has been analyzed.

Россия, г. Пенза, Пензенский государственный технологический университет Currently, progress does not stand still and the industries are constantly developing. Education is one of them. New teaching methods and systems are emerging all the time, all for the sake of one goal -to educate people in society, to make it more intellectual, for the sake of people's development. The article discusses the introduction of distance learning services, the program of which will automatically adapt to a specific individual, increasing the effectiveness of studying the course at times. For effective and efficient work, the technology of artificial neural networks is used, with the possibility of dynamic learning. A feed-forward network is selected. The neural network is trained using the error backpropagation method. The prototype proposed in this article can serve as a starting point for the creation of an improved version of distance learning that can develop the speed of understanding and learning of the courses of the learners.