“We Read, We Tweet” geographically visualizes the dissemination of New York Times articles through Twitter. Each line connects the location of a tweet to the contextual location of the New York Times article it referenced. The lines are generated in a sequence based on the time in which a tweet occurs. The project explores digital news distribution in a temporal and spatial context through the social space of Twitter. The video only shows a small portion of tweets aggregated from each article, some of which contain of corresponding Tweets.
The articles and tweets are constantly being aggregated and stored in a database, making use of the Twitter, Backtweets, Google Maps, and New York Times Articles API. Every 10 minutes, the Backtweets API is queried to find the most recent New York Times articles that have been tweeted about. For each article found, the New York Times Articles API is queried and if a contextual location is found, that location is then geocoded using the Google Maps API. Every tweet that mentions this article is also geocoded using the Google Maps API, and both the article and tweets are stored in a database. The Backtweets API was quite awesome with parsing out shortened URLs. The visualization itself was written in Java / Processing.
