Dataset for Human-AI Synthesis of Divergent Crisis Narratives: BBC Persian and Tasnim News Coverage of the 2026 Iran Conflict

This dataset accompanies the research study entitled: "Human-AI Synthesis of Divergent Crisis Narratives: Evidence from the 2026 Iran Conflict". The dataset contains 21,304 Persian-language news reports collected from the official Telegram channels of BBC Persian and Tasnim News Agency during the 2026 Iran Conflict. The repository includes: • Full dataset of collected reports • Crisis-related subset • Topic modeling outputs • Semantic event alignment results • Narrative divergence analysis results • Conflict and discrepancy datasets • Human validation records • Event comparison outputs • Summary statistics The dataset was created to support research in Human-AI collaboration, Cognitive Infocommunications (CogInfoCom), crisis knowledge management, narrative divergence analysis, event alignment, and trustworthy knowledge synthesis. Key findings derived from the dataset include: • Coverage Ratio = 30.5% • Narrative Divergence Index (NDI) = 0.695 • Average Event Similarity = 0.755 • Highly Aligned Events = 21 Researchers may use this dataset for replication studies, benchmarking, narrative analysis, event alignment, media comparison, crisis communication studies, and Human-AI collaboration research.