Visualizing Social Media Content with SentenTree

We introduce SentenTree, a novel technique for visualizing the content of unstructured social media text. SentenTree displays frequent sentence patterns abstracted from a corpus of social media posts. The technique employs design ideas from word clouds and the Word Tree, but overcomes a number of limitations of both those visualizations. SentenTree displays a node-link diagram where nodes are words and links indicate word co-occurrence within the same sentence. The spatial arrangement of nodes gives cues to the syntactic ordering of words while the size of nodes gives cues to their frequency of occurrence. SentenTree can help people gain a rapid understanding of key concepts and opinions in a large social media text collection. It is implemented as a lightweight application that runs in the browser.