Artificial Intelligence-Based Linguopragmatic Analysis of Multimodal Communication in Uzbek Telegram Social Media
Year: 2026
Type: conference-paper
Abstract: The rapid expansion of social media has fundamentally transformed the structure of digital communication by integrating verbal and non-verbal semiotic resources into unified multimodal discourse. Among contemporary communication platforms, Telegram has become one of the most influential environments for Uzbek-language interaction, where written texts are continuously accompanied by emojis, stickers, images, voice messages, hyperlinks, GIF animations, and other visual elements. These multimodal units function not merely as complementary signs but as linguopragmatic mechanisms that shape communicative intention, interpersonal stance, emotional evaluation, and contextual interpretation. Despite the growing international interest in multimodal discourse and artificial intelligence-assisted language processing, computational investigations of Uzbek multimodal communication remain limited due to the scarcity of annotated corpora and low-resource language technologies. This study proposes an artificial intelligence-based framework for linguopragmatic analysis of multimodal communication in Uzbek Telegram social media. A multimodal corpus of Uzbek Telegram messages is employed to identify recurrent multimodal structures and to classify their communicative functions through AI-assisted annotation supported by natural language processing techniques. The research combines discourse analysis, pragmalinguistics, multimodal semiotics, corpus linguistics, and computational methods to reveal how textual and visual resources jointly construct pragmatic meaning. The findings demonstrate that multimodal units substantially enhance communicative efficiency by strengthening evaluation, presupposition, emotional expressiveness, implicature, and persuasive interaction. Furthermore, the proposed analytical framework contributes to the development of computational linguistics and multimodal NLP resources for the Uzbek language while expanding theoretical understanding of digital pragmatics in low-resource languages
Language: English
Subfield: Linguistics and Language
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7224770367