This document discusses using spatial analysis and mixture of Gaussians modeling to analyze geo-tagged tweets from a city to identify hot spots and patterns in people's behavior over time and location. The goal is to empirically model the spatial density of tweets. The document describes using expectation maximization to fit the mixture of Gaussians model to synthetic and real tweet location data, and issues that arose such as model collapsing and the lack of a global maximum. BIC was used to select the number of clusters but also had limitations. Future work proposed focusing on city centers and understanding when BIC works best.