Florida International University
Joshua Daniel Eisenberg is a Ph.D. candidate under the direction of Mark Finlayson in the Cognac Laboratory. In 2012 he earned B.A. in mathematics, and a minor in computer science from Brandeis University. In 2014 he earned a B.S. in computer engineering from Florida International University. He was a fellow for the NSF Open Science Data Cloud PIRE fellowship in 2013, 2014, and 2016. Outside of academics, Joshua is interested in improvisational music, music festivals, plants, traveling, art, and food.
My dissertation is on the automatic extraction of narrative structure from English text. I am developing programs that allow computers to extract structural information from stories: given a text, the computer will detect if there are stories in it, where they occur (narrative boundaries), what is the plot of each story, and other information about the narrator. These are aspects of narrative that people are aware of when listening to a story, and we use this information to interpret the story. Currently computers do not have these abilities.
Story understanding is an almost an automatic awareness for people, but only because we’ve been trained to process natural language into stories. Even though people can easily understand stories, their structure is not always so simple. People don’t tell one story at a time. We don’t specifically announce when we are starting or stopping our stories. We interrupt each other. We tell stories within stories. As of yet, computers have no machinery that allows them to transform a story in natural language into understanding of what actually happened in the story, and use this to help them make informed decisions based on what happened. In this presentation I will discuss the ways I have enabled computers to understand narrative, and my plans for the remainder of my dissertation research.