Twitter-Based Geocollaboration: Geovisualization and Geotagging of Microblogging Messages

Abstract

This chapter describes a web-based model of geocollaboration based on geolocalized tweets. From the Twitter stream, it allows selecting those geotagged messages, filtering them by content, and placing them on a map. This map can be zoomed at street level in order to read the displayed messages, or zoomed out to visualize their geographic distribution and frequency. This approach also allows posting messages on the map with an associated geolocalization, other than the user’s current location, corresponding to a specific location where an event is reported or information is requested. Furthermore, all displayed messages can be replied on the map, thus creating a geolocalized conversation thread to support collaboration. Accordingly, our model addresses four main problems: collecting geotagged tweets, geovisualizing these messages, posting and geolocalizing new messages, and supporting geolocalized conversation threading. Experiments showing the promise of the approach to deal with web-based communication issues in catastrophe situations, such as an earthquake or flood, are also described.