TweetPos: A Tool to Study the Geographic Evolution of Twitter Topics

Abstract

Popular Social Networking Sites (SNSs) like Twitter and Facebook are evolving into crowd-sourced, interdisciplinary sensor systems that “monitor” a wide spectrum of (physical) properties and topics. This paper introduces TweetPos, a web service that is intended to facilitate the analytical study of geographic tendencies in Twitter data feeds. To oblige the human cognitive features, the TweetPos tool maximally relies on visual data structures like heatmaps and charts to represent the geo-spatial sources of tweets. The tool compiles data bodies that grant insight in both past and present tweet posting behavior, incorporates an animation engine to highlight temporal trends, and leverages layered visualization techniques so that multiple topics can be offset against each other, all from a geographic perspective. Via the presentation of two representative use cases, we comprehensively demonstrate TweetPos’ data mining and analytical features and we illustrate the (geo-spatial) intelligence they can amount to. Thanks to a generic implementation, the TweetPos service is not geared towards a specific target audience but instead is sufficiently versatile to be valuable for a vast and varied collection of consumer profiles like social scientists and market analysts.