Presentation Information
[B-7-04]Crisis Early Warning via Low-Frequency Search Trend Analysis: The 2026 Iran Crisis
◎△YU CHEN1, Kazuki Nakajima1, Masaki Aida1 (1. Tokyo Metropolitan University)
Keywords:
nline User Dynamics,Low-Frequency Mode Analysis,Search Trend Analysis,Early Warning,Google Trends
This study investigates whether low-frequency mode analysis of Google Trends search data can capture early changes in online search behavior related to international crises. Previous studies on online user dynamics have shown that low-frequency modes may emerge in activity time series before excessive activation of online users. As a case study, we analyze the 2026 Iran crisis using the search keyword “Iran.” A sliding-window discrete Fourier transform is applied to the Google Trends time series from December 2025 to May 24, 2026. After removing the DC component and normalizing the amplitude spectrum, the lowest-frequency component is extracted at each time point. The results show that, although the search frequency sharply increased around March 1, 2026, the lowest-frequency component had already increased from mid-January to early February. This suggests that low-frequency mode analysis may capture changes in the temporal structure of search behavior before they become visible as a large increase in search volume.