The impact of social media on consumer choice

The formation of a modern marketing strategy in the digital environment involves the use of various data that make it possible to assess the current situation and identify development pros­pects. The hypothesis is formulated that sentiment and semantic analysis using machine learning algorithms allows businesses to objec­tively assess the attitude of the target audience to the activities of brands on the Internet and identify popular thematic content. Conducting the research, general scientific methods of analysis and synthesis were used to characterize the basic principles of using sentiment and semantic analysis in the process of assessing brand reputation; empirical methods, graphic representation, and system-structural analysis. The feasibility of using text information and emoticons as a valuable source of data for developing effective manage­ment decisions in the field of marketing is proven. The implementation of sentiment and semantic analysis based on text and emoji is justified. A structural and logical scheme of the differences between sentiment and semantic analysis is presented. The features of building information support when implementing the two approaches under study are investigated. The necessity of using social media for assembling text content and emoticons is proven, which is asso­cia­ted with the significant activity of generations Y, Z, and Alpha on such platforms as YouTube, Instagram, TikTok, etc. The feasibility of using sentiment and semantic analysis for retail chains that sell consumer electronics on the Internet is proven. Machine learning algorithms are used to assess user sentiment and interests, which allows for effective processing of text content