| news . people . research . publications . teaching . software . data . lab . contact |
The goal of this project is to create a visualization system beyond the simple highlighting of raw text that can provide awareness of the kinds of features extracted and the relative volume of those features in large amounts of unstructured text (could have different answers for emails, tweets, research papers, mixed jumble of hard drive folders, etc.). Features could encompass n-grams, extracted entity mentions, or uber-concepts aligned to an ontology. Some aspects of particular interest - who is communicating in these texts? How does conceptual content of communications drift over time? In asynchronous vs synchronous communications? How to represent the external referents of the texts?
Presentation
