Hidden Persuasion: Detecting Manipulative Narratives on Social Media During the 2022 Russian Invasion of Ukraine

TL;DR

This paper introduces a high-performing method for detecting manipulative narratives on social media, specifically targeting Ukrainian Telegram during the 2022 Russian invasion, using fine-tuned language models and advanced classification techniques.

Contribution

It presents a novel approach combining fine-tuned language models with meta-feature classifiers for manipulation detection in social media content.

Findings

Achieved 2nd place in classification accuracy

Achieved 3rd place in span detection performance

Demonstrated effectiveness of fine-tuned language models for manipulation detection

Abstract

This paper presents one of the top-performing solutions to the UNLP 2025 Shared Task on Detecting Manipulation in Social Media. The task focuses on detecting and classifying rhetorical and stylistic manipulation techniques used to influence Ukrainian Telegram users. For the classification subtask, we fine-tuned the Gemma 2 language model with LoRA adapters and applied a second-level classifier leveraging meta-features and threshold optimization. For span detection, we employed an XLM-RoBERTa model trained for multi-target, including token binary classification. Our approach achieved 2nd place in classification and 3rd place in span detection.