Why Social Media Engagement Is Not Public Opinion My new article in Fair Observer examines how a small minority of highly active users can dominate visibility online and create the illusion of majority opinion. Read: www.fairobserver.com/world-news/m...

Media coverage highlights conflict. Public interest points elsewhere. Search demand was driven more by economic pressure, daily living concerns, and emigration. Full analysis + data: www.soufan.tech/analysis/leb...

Beirut Strikes Media Coverage Bias To measure how global media framed the event, I analyzed English-language news headlines collected via the GDELT Project in the 48 hours following the strikes (April 8–9, 2026). @eurasiareview.bsky.social

After Beirut strikes, 73.5% of media coverage focused on politics, not civilians. Only 26.5% covered human impact. Displacement appeared in just 1.8%. www.soufan.tech/analysis/leb...

#Lebanon strike reports surged 5-fold before the #ceasefire. Now #Beirut strikes align with that signal. Data was early: www.soufan.tech/analysis/lebanon-strike-reports-5x-before-ceasefire-x/

Strike reports in south Lebanon surged to 5× normal levels before the ceasefire and stayed elevated as it was announced. Based on 5,285 posts on X. Full analysis: www.soufan.tech/analysis/leb...

Ain Saadeh incident analysis: Political posts earned 4.6× more engagement per post than strike claims, despite making up only 5% of the conversation. www.soufan.tech/analysis/ain...

A study analyzing 16,695 Arabic-language tweets finds that posts expressing linguistic uncertainty receive significantly higher engagement, particularly replies—an effect described as “uncertainty–reply asymmetry.”

Posts expressing uncertainty receive higher engagement across all interaction types, with likes about 36.4% higher, reposts 33.3% higher, and replies 50.0% higher than certainty-framed tweets. www.soufan.tech/data-briefs/...

My Python data pipeline for analyzing 16,695 Arabic tweets and studying linguistic uncertainty and engagement patterns on X is now published on HackerNoon. Part of my research on Uncertainty–Reply Asymmetry.