Research Papers

Research papers by Mohamed Soufan on Lebanon, media, AI, digital behavior and computational social research, with links to full papers, summaries and code.

Explore research papers by Mohamed Soufan on Lebanon, media, artificial intelligence, digital behavior and computational social research. Each entry links to the full paper and, where available, a research summary, code and data.


SSRN · 2026

Analyzes 142 fact-checked misinformation claims related to Lebanon’s 2026 conflict and finds that 85.9% involved authentic content used deceptively, compared with 11.3% classified as AI-related. Only 1.4% of claims were identified in professional news coverage under the study’s search protocol.

Full paper · Research summary · Code & data


Language-dependent Source Representation in AI Search: A Cross-platform Audit of Lebanon

SSRN · 2026

Analyzes 320 responses across ChatGPT Search, Google Gemini, Perplexity, and Microsoft Copilot, finding that Arabic-language queries surfaced Lebanese-origin sources substantially more often than matched English-language queries, with an estimated difference of 22.9 percentage points across questions.

Full paper · Research summary


Geographic Disparities in International News Coverage of Fatal Israeli Attacks in Lebanon, March 2026

SSRN · 2026

Analyzes 229 fatal location-days during Lebanon’s March 2026 escalation and finds that South Lebanon and Nabatieh had about 65% lower adjusted odds of receiving international English-language news coverage than fatal events elsewhere in the country.

Full paper · Research summary · Code & data


Online Remembrance of the Beirut Port Explosion: Justice, State Failure, and Digital Memory on X and Reddit

SSRN · 2026

Analyzes 7,191 Arabic- and English-language posts and comments about the Beirut port explosion, finding that unresolved justice and state failure accounted for 71.7% of online memory around the five-year mark, far outweighing personal grief.

Full paper · Research summary · Code & data


Visibility Without Guidance: Measuring the Actionability of Public Crisis Communication in Lebanon

arXiv · 2026

Audits 182 public crisis-communication items from nine Lebanese state, emergency and UN sources, finding a mean actionability score of just 1.98 out of 5 and no access or contact information in 89.6% of items.

Full paper · Research summary · Code & data


Linguistic Uncertainty and Reply Engagement on X: A Cross-Domain Replication of the Uncertainty-Reply Asymmetry

arXiv · 2026

Analyzes 2,258 English-language posts on Federal Reserve policy, inflation, and electoral politics and identifies an uncertainty-reply asymmetry: posts expressing linguistic uncertainty received 82% more replies on average and approximately 13% higher expected reply engagement after controls.

Full paper · Research summary


Measuring the Gap Between Media Coverage and Public Information Demand: Evidence from the 2026 Lebanon Conflict

arXiv · 2026

Analyzes 11,623 English-language news articles and Google Trends data during the March 2026 Lebanon conflict, finding that conflict dominated 94.9% of media coverage while accounting for only 36.9% of public search interest.

Full paper · Research summary · Code & data


The Structure of Participation and Attention in Arabic-Language Hezbollah Discourse on X

arXiv · 2026

Analyzes 15,767 Arabic-language posts about Hezbollah on X and finds that audience attention was highly concentrated, with the top 1% of users capturing 61.5% of all engagement despite broad participation.

Full paper · Research summary


Linguistic Uncertainty and Engagement in Arabic-Language X (formerly Twitter) Discourse

arXiv · 2026

Analyzes 16,695 Arabic-language tweets about Lebanon and identifies an uncertainty-reply asymmetry: posts expressing linguistic uncertainty received approximately 25% higher expected engagement, with the strongest association observed for replies.

Full paper · Research summary