Propaganda analysis

Abstract

The pervasive influence of propaganda across various media platforms poses significant chal lenges for the discernment of biased information. This project report presents a comprehensive examination of propaganda, proposing a nuanced definition that encapsulates its multifaceted nature, which includes content, intent, techniques, medium, and impact among other elements. We explore the challenges associated with collecting data from diverse sources like social media platforms and news outlets, and describe our pivot to sourcing content through tailored search queries to compile a dataset of propagandistic articles. This dataset aims to enable sophisti cated Natural Language Processing (NLP) models to identify and analyze the subtle nuances of propaganda. Additionally, we detail our rigorous annotation process, which includes guidelines for identifying propaganda techniques and narrative classes, and outline measures to ensure high inter-annotator reliability. By enhancing the understanding of propaganda’s complex dynamics, this work contributes to the development of more effective tools for media literacy and informa tion integrity in the digital age.