Manipulative technologies in a large amount of linguistic data: Social media

Abstract

The article presents the results of our study on manipulative techniques based on a large amount of linguistic data from social media and focused on the implementation of a major urban development project in the field of road transport construction in Moscow. The research was conducted using data from social networks, video hosting platforms, microblogs, blogs, forums and review websites (3,358,590 tokens, the audience of 17,688,221). Created during the study, the database contains user-generated content and their digital footprints reflecting user reactions. The data analysis was conducted from a cognitive perspective, employing a neural network text analysis, content analysis, sentiment analysis, digital aggression analysis and lexical association analysis. The Brand Analytics monitoring system was used for data collection, while GPT-4o, GPT-4o mini, GPT-4, TextAnalyst 2.32 and AutoMap were used for the data analysis and interpretation, with the Tableau platform used for visual analytics. The analysis of the data revealed mythological manipulation, speech and psychotechnologies. When manipulative techniques are employed in the media space, rhetorical tropes and figures, sophisms and eristic tricks are actively used. The adaptation of these techniques to the specifics of digital communication allows for indirect influence on the audience, significantly increasing the effectiveness of the impact, which can, in turn, lead to heightened conflict potential and social tension.