Gauging voter turnout and voter behavior is difficult to do, yet can have consequences for politicians and public on what voter issues are most significant at any given time.


Tweets, hashtags, searcy query data and even what type of car an individual owns can all be used to predict voting preference and even polling outcomes.


UCIPT researchers are looking at how this social data can be used to predict and highlight areas of high vulnerability for politicians. We are also developing algorithms that can gauge the relative importance of political issues before they hit the public.