Dataset of Propaganda and Non-Propaganda Poster Titles

The dataset consists of titles from propaganda and non-propaganda posters, categorized for binary classification research. These titles capture textual elements from materials designed to influence or convey messages, focusing on distinguishing propaganda from non-propaganda content. Labeled with binary tags (propaganda/non-propaganda), the dataset supports studies on text classification, exploring linguistic and stylistic differences. Sourced and labeled for clarity and real-world relevance, it enables comparative analysis of linguistic trends, stylistic features, and propaganda's role in media.

This is raw text data, requiring preprocessing steps such as data cleaning, removing special characters, and other standard text preparation techniques.
Two folders containing sample poster images are also included: one for propaganda posters and the other for non-propaganda posters.
This resource is valuable for researchers in computer science, social sciences, and media studies investigating automated detection, language analysis, and public perception of propaganda.

本数据集由宣传与非宣传海报的标题构成，旨在为二元分类研究提供分类。这些标题提炼自旨在影响或传达信息的材料，专注于区分宣传内容与非宣传内容。数据集经二元标签（宣传/非宣传）标注，支持文本分类研究，探索语言和风格的差异。数据集的收集与标注旨在清晰度与现实相关性，从而促进语言趋势、风格特征与宣传在媒体中作用的比较分析。此为原始文本数据，需经过数据清洗、去除特殊字符等标准文本准备步骤。此外，还包含两个包含样本海报图像的文件夹：一个用于宣传海报，另一个用于非宣传海报。对于研究计算机科学、社会科学和媒体学的学者而言，该资源对于研究自动检测、语言分析和公众对宣传的认知具有重要价值。