Title: Automatic Text Summarization

Abstract:
Automatic text summarization, the generation of a condensed version of a text by a computer, has been a focus of research for several decades. With the overwhelming amount of information available on the web, summarization is now becoming an acutely needed technology. 
In this talk, I will give an overview of the challenges entailed in building an automatic summarizer, and describe the consensus architecture of such systems.
I will then focus on a sub-problem of summarization: information ordering, i.e., how to find an optimal order in which to present the information to be conveyed in a summary. I will describe our methodology, our proposed solution, and the results of our evaluation studies conducted with human subjects.

Bio:
Noemie Elhadad is an assistant professor in the CS department at CCNY. Her research interests are in natural language processing (text summarization, statistical text generation and user modeling). She completed her PhD from Columbia University in 2005. This semester she teaches "Statistical Natural Language Processing" at the graduate center.