Using Search Queries to Analyze Public Attention to One-Time Political Events

The paper introduces a simple empirical method to estimate the dynamics of the number of search requests over several weeks after one-time political events, such as referendum, one-round elections or coups d'etat. We apply the method to three events in 2016, namely the U.S. Presidential election, Brexit and the attempted coup d'etat in Turkey. The empirical method uses Google Trends as a source of raw data. A mathematical model is also proposed that fits empirical data. The model is based on Rashevsky-style decision-making scheme.