A Social Media Analysis of Discourse on the Israel--Palestine Conflict on Telegram

A research study conducted a multi-method computational analysis of discourse on the Israel-Palestine conflict on Telegram, comparing pro-Israel and pro-Palestine communities. The study analyzed 87,617 messages from sixteen channels between May 2021 and June 2026, employing sentiment analysis, three stance detection methods, and framing analysis. The core objective was to systematically compare these communities at scale, providing insight into deliberate political communication within the platform's broadcast architecture.

Why it matters

This research is strategically important because it sheds light on how digital platforms are used to shape narratives and influence public opinion during conflicts. Understanding the dynamics of such discourse on platforms like Telegram can inform strategies for managing information environments and countering disinformation campaigns in sensitive geopolitical contexts.

Key insights

• Social media platforms, specifically Telegram, serve as a central arena for contesting armed conflicts.
• Pro-Israel and pro-Palestine communities on Telegram, despite its broadcast architecture, had not been systematically compared at scale prior to this study.
• The research utilized a multi-method computational analysis approach.
• The dataset comprised 87,617 messages from sixteen Telegram channels (eight pro-Israel, eight pro-Palestine).
• The analysis covered the period from May 2021 to June 2026, encompassing multiple conflict escalations.
• Methods employed included sentiment analysis, three distinct stance detection methods (keyword matching, zero-shot DeBERTa, fine-tuned BERTweet), and framing analysis.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2608.21385