| Abstract | Disinformation remains one of the major challenges in today's digital landscape. Although numerous detection approaches have been proposed, the vast volume and dynamic nature of social media data, combined with privacy and legal constraints, continue to hinder research and mitigation efforts. In this work, we present a dataset of 977 publicly accessible Telegram channels and groups relevant to disinformation in Germany, collected over more than one year, encompassing significant events such as the war in Ukraine and four state elections in Germany. To address privacy and legal concerns, we apply cleaning and pseudonymization processes while preserving structural and behavioral details. Our dataset enables metadatabased rather than content-based analysis, allowing efficient, lowresource exploration of dissemination patterns without requiring deep technical expertise or platform-specific knowledge. The dataset is designed to support broader network analysis of disinformation dynamics within Telegram's unique communication ecosystem while safeguarding privacy. |
|---|
