Skip to content
Conference Open access

A Computational-Pragmatic Framework for Nonverbal Units in Digital Discourse

Sep 2026 · 2026 11th International Conference on Computer Science and Engineering (UBMK) · pp. 1562-1567 · 0 citations · 19 references

Abstract

With the fast-paced advancement of digital communication, there is a rising trend of using nonverbal units like emojis, stickers, animated images, reactions, and typography in online communication. Such multimodal units are widely used to express emotions, make evaluations, show interpersonal attitudes, and manage conversations. Nevertheless, their pragmatic functions have not been studied systematically enough through the computational perspective. In this paper, a computational-pragmatic approach to modeling nonverbal units in digital discourse is proposed. A pilot corpus of 500 nonverbal units was collected from Telegram, Instagram, Facebook, and YouTube and analyzed. To systematically annotate nonverbal pragmatic units, the Nonverbal Pragmatic Tagset was developed, including such categories as emotional expression, evaluation, politeness, discourse management, emphasis, humor, irony, and sarcasm. The analysis involved three main components: pragmatic annotation, sentiment analysis, and exploratory semantic similarity analysis based on the Sentence-Bidirectional Encoder Representations from Transformers model. The analysis showed that the most frequently occurring pragmatic functions were emotional expression (29.0%) and evaluation (19.6%), and the predominant sentiment in the corpus was positive (49.6%). Moreover, exploratory semantic analysis using the Sentence-Bidirectional Encoder Representations from Transformers model revealed broad semantic tendencies that were generally compatible with the manually assigned pragmatic tags and sentiment polarity. Thus, the proposed computational-pragmatic framework provides a structured approach for describing the pragmatic, emotional, and semantic properties of nonverbal units in digital discourse.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.