Questions and features of automated social networks monitoring with support of intelligent user messages analysis.

Вестник Брянского государственного технического университета. 2015. № 3(47) 117 УДК 004.85 Д.В. Будыльский, А.Г. Подвесовский ИССЛЕДОВАНИЕ ПРИМЕНИМОСТИ МОДЕЛЕЙ ГЛУБОКОГО ОБУЧЕНИЯ ДЛЯ ЗАДАЧИ АСПЕКТНОГО АНАЛИЗА ТОНАЛЬНОСТИ ТЕКСТОВЫХ СООБЩЕНИЙРассмотрена задача аспектного анализа тональности текстовых сообщений на естественном языке. Исследо-ваны четыре нейросетевые модели, относящиеся к разделу глубокого обучения: сверточная нейронная сеть, рекуррентная нейронная сеть, сеть GRU, сеть LSTM. Представлены результаты экспериментальной провер-ки указанных моделей на корпусе текстовых отзывов SentiRuEval-2015.Ключевые слова: машинное обучение, аспектный анализ тональности, нейронные сети, глубокое обучение.

show abstract

The methodology and methods of sociological monitoring of communities of working youth in social networks

Malykhina

Bocharov

2022

Semiotic studies

View full text Add to dashboard Cite

The article is devoted to the generalization of the experience accumulated by our research group in the research of virtual communities of the working youth (2019-2021). The purpose of the article is to construct a comprehensive methodology for the socio-logical monitoring of social networks of the working youth. The main research objectives were: to analyze modern approaches to conducting monitoring research (including on the Internet); based on the methodology of discourse analysis, to consider the use of methods of frequent and infrequent content analysis for processing the collected information in monitoring mode; to present the stages of monitoring communities of the working youth in social networks, using the example of research conducted within the framework of the project Life strategies of young people referring to the new working class in modern Russia. As a part of the empirical data analysis, the results of a study of virtual communities and social networks groups of the Russian working youth in the field of customer service in the VKontakte social network, carried out in 2021 (6 virtual communities, 133 text posts and the most popular memes) are presented in detail. The stability of thematic categories of virtual group communication in comparison with previous studies and the presence of group identity among Russian working youth referring to the custom service sphere. Conclusions are made regarding the preference of a non-automated approach to the collection and analysis of social network content; about the validity of using the methodology of discourse analysis and content analysis methods for the research of youth social networks in the mode of sociological monitoring.

show abstract

Questions and features of automated social networks monitoring with support of intelligent user messages analysis.

Cited by 3 publications

References 1 publication

Development of a Complex of Algorithms for Collecting and Marking Text Data for Identifying Destructive Messages

Development of a Complex of Algorithms for Collecting and Marking Text Data for Identifying Destructive Messages

Application of deep learning models for aspect based sentiment analysis.

The methodology and methods of sociological monitoring of communities of working youth in social networks

Contact Info

Product

Resources

About