Social Media usage is becoming more and more interwoven with activities in urban space. Understanding the flows of information through the cities can open new doors for us to understand how urban space relates to human behavior. In this paper, we introduce a method to extrapolate flow of geolocated social media data on a street network. This method allows for combination of flow and semantic data from social media. We then apply this method to a corpus of geolocated tweets collected from the Los Angeles metropolitan area. We compared the results to betweenness centrality of the streets as a measurement of connectivity and density of businesses as a measurement of public activity. We find that the flows calculated from Twitter have a higher correlation with public activities hinting towards the relationship between geolocated social media usage and businesses and public space.