Propaganda Content Identification

Identifying propaganda in online content is a crucial research area aiming to mitigate the spread of misinformation and manipulative narratives. Current efforts focus on developing automated systems, often leveraging large language models (LLMs), to detect propaganda techniques, appeals, and underlying intent at both sentence and token levels, incorporating discourse structure analysis to improve accuracy. These advancements rely heavily on the creation and utilization of high-quality multilingual datasets, improving the ability to identify propaganda across diverse languages and contexts, with a particular focus on improving the performance of LLMs in this task. This research has significant implications for combating disinformation and promoting informed public discourse.