Daria's profile picture

Daria Sokova


Leverhulme Trust Doctoral Scholar

Academic and research departments

Literature and Languages.

About

My research project

Publications

Anastasiia Bezobrazova, Daria Sokova, Constantin Orasan (2026)Emotion-aware text simplification of user generated content using LLMs, In: The Proceedings for the 15th Workshop on Computational Approaches to Subjectivity, Sentiment Social Media Analysis (WASSA 2026)pp. 107-122 Association for Computational Linguistics

Digital inclusion increasingly supports adults with intellectual disabilities (ID) to participate online, yet social media posts can be difficult to understand, particularly when they contain strong emotions, slang, or non-standard writing. This paper investigates whether large language models (LLMs) can simplify social media texts to improve cognitive accessibility and preserve emotional meaning. Using an accessibility-oriented prompt based on existing guidance, posts are simplified and emotion preservation is assessed. The results suggest that many simplified posts retain the same emotions, though changes occur, especially when emotions are weakly expressed or ambiguous. Qualitative analysis shows that simplification improves fluency and structure but can also shift perceived emotion through changes to tone, formatting, and other affective cues common in social media text. The research has also revealed that different LLMs produce very different outputs.

Daria Sokova, Anastasiia Bezobrazova, Constantin Orasan (2025)SQUREL at TSAR 2025 Shared Task: CEFR-Controlled Text Simplification with Prompting and Reinforcement Fine-Tuning, In: Proceedings of the Fourth Workshop on Text Simplification, Accessibility and Readability (TSAR 2025)pp. 242-250 Association for Computational Linguistics

This paper summarises the submissions of our team to the TSAR 2025 Shared Task on Readability-Controlled Text Simplification, which aims to create text simplifications that balance reduced linguistic complexity, meaning preservation, and fluency while meeting a predefined target readability level. In this work, we proposed two different methods for CEFR-controlled text simplification: a setup which employed reinforcement fine-tuning of large language models (LLMs) and a conservative lexical pipeline which relied on prompting LLMs to simplify sentences.