To overcome this location sparsity problem, we propose in this paper to predict a user’s location based purely on the content of the user’s tweets, even in the absence of any other geospatial cues. Our intuition is that a user’s tweets may encode some location-specific content – either specific place names or certain words or phrases more likely to be associated with certain locations than others (e.g., “howdy” for people from Texas). In this way, we can fill-the-gap for the 74% of Twitter users lacking city-level granular location information. By augmenting the massive human-powered sensing capabilities of Twitter and related microblogging services with content-derived location information, this framework can overcome the sparsity of geo-enabled features in these services and bring augmented scope and breadth to emerging location-based personalized information services.