Timeline visualization of semantic content

Abstract

People interact with large corpuses of documents everyday, from Googling the internet, reading a book, or checking up on their email. Much of this content has a temporal component: a Website was published on a particular date, your email arrived yesterday, and Chapter 2 comes after Chapter 1. As we read this content, we create an internal map that correlates what we read with its place in time and with other parts that we've read. The quality of this map is critical to understanding the structure of any large corpus and for locating salient information.