Detecting persuasive techniques in Ukrainian wartime memes with AI
UkrMeme: Multimodal Explainable Persuasion Detection in Ukrainian Memes
Artificial IntelligenceComputation and LanguageComputer Vision and Pattern Recognition
Summary
Memes mix pictures and words and can persuade people, especially during conflicts. The authors created a new collection of Ukrainian wartime memes labeled with the ways they try to persuade viewers. They tested AI models that understand both images and text to spot these persuasive techniques without extra training. They also compared different methods for explaining how the AI makes these decisions, checking which parts of the meme were most important.
What this means in practice
- •For social media moderation teams: Automatically identify persuasive or propaganda techniques in memes to assist content moderation and flagging in conflict-related contexts.
- •For digital marketing analysts: Analyze how persuasive visual and textual elements combine in online memes to better understand audience influence strategies.
Authors
Nazar Andrushko, Vera Schmitt, Veronika Solopova
Abstract
Memes play a central role in modern online communication, especially in contexts of political conflict where visual and textual cues combine to convey persuasive messages. We study persuasive technique detection in Ukrainian memes using vision-language models (VLMs). We introduce UkrMeme, a dataset of 200 organically produced Ukrainian wartime memes annotated for persuasion techniques, and PTM-UA, a Ukrainian translation of a subset of the Propaganda 950 benchmark that preserves original visual content while translating meme text. We evaluate VLMs in a zero-shot prompting setup across image, text, and multimodal inputs using hierarchical F1 metrics. Finally, we compare Layer-wise Relevance Propagation (LRP) and model-generated token attributions and analyze the semantic categories of evidence emphasized by each explanation method.