NewsScatter: Topic Similarity in Social Media

Abstract

News organisations that use social media sites (such as Twitter) rapidly generate a large volume of data every day. Data visualisation is an effective way to represent microblogging data graphically in order to increase understanding of what news organisations are reporting over a given time period. This paper focuses on visualising the tweets (posts on Twitter) of a broad cross-section of English-language news outlets over time, with the goal of developing an interactive web-based visualisation system that summarises the output of news outlets’ tweets as a scatter plot where each point refers to a news outlet. Such a visualisation could enable interested parties to explore similarities and differences among what news outlets report based on the frequency with which specific words are used.