Thematic modeling and linguistic analysis of text messages from a social network for information and analytical support of logistics business

Abstract

In the modern economy, the success of a business is largely determined by the company’s ability to analyze consumer preferences, consumer attitudes towards the company’s products, as well as the ability to quickly respond to changing preferences or negative trends. Social listening is a technology for analyzing conversations, text messages and any kind of mention of a company, its products or brand. Currently, it is most effective to carry out social listening by monitoring social networks (VKontakte, etc.), which are the largest sources of text messages from millions of users. The purpose of this work is to analyze the practices of using social listening technology, as well as common approaches to the use of social networks by domestic and foreign companies. Based on the specialized software developed by the authors, an analysis of more than 50 000 news reports published in 2021–2024 was carried out on companies of different levels and specialization. Using linguistic analysis of the corpus of text messages for various companies and sectors of the economy, the most common words were identified, thematic modeling was carried out, and the dynamics of news reports and their relationship with external factors were studied.