用AI分析阿拉伯语网络内容，发现六大类阴谋论
Uncovering Conspiratorial Narratives within Arabic Online Content
- 结合命名实体识别与Top2Vec算法分析阿拉伯博客和脸书数据
- 识别出性别、地缘政治、政府掩盖等六类阴谋叙事
- 填补了阿拉伯语网络阴谋论研究的空白，适合社科与舆情研究者
本研究通过计算方法分析阿拉伯语数字空间中的在线内容，揭示阴谋论的传播模式。采用命名实体识别与主题建模（特别是Top2Vec算法），对阿拉伯语博客和脸书数据进行分析，识别并分类出六类阴谋叙事：性别/女权、地缘政治、政府掩盖、末世论、犹太-共济会、地球工程。研究指出这些叙事深深嵌入阿拉伯社交媒体话语中，受地区历史、文化与社会政治背景影响。通过应用先进的自然语言处理技术分析阿拉伯语内容，本研究弥补了传统阴谋论研究多集中于英语或线下数据的不足。结果为理解阿拉伯语数字空间中阴谋论的表现形式与演变提供了新视角，深化了对其在阿拉伯世界公共舆论塑造中作用的认识。
原文摘要 · Abstract (English)
This study investigates the spread of conspiracy theories in Arabic digital spaces through computational analysis of online content. By combining Named Entity Recognition and Topic Modeling techniques, specifically the Top2Vec algorithm, we analyze data from Arabic blogs and Facebook to identify and classify conspiratorial narratives. Our analysis uncovers six distinct categories: gender/feminist, geopolitical, government cover-ups, apocalyptic, Judeo-Masonic, and geoengineering. The research highlights how these narratives are deeply embedded in Arabic social media discourse, shaped by regional historical, cultural, and sociopolitical contexts. By applying advanced Natural Language Processing methods to Arabic content, this study addresses a gap in conspiracy theory research, which has traditionally focused on English-language content or offline data. The findings provide new insights into the manifestation and evolution of conspiracy theories in Arabic digital spaces, enhancing our understanding of their role in shaping public discourse in the Arab world.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。