The Computer Science Colloquium

Noemie Elhadad
(CCNY)

"User-Sensitive Text Summarization"

Automatic text summarization has become increasingly established over the past ten years. However, most summarizers to date produce generic summaries of text, independent of the reader. In this talk, I present a novel approach to text summarization that generates user-sensitive text summaries. I present the design and implementation of a summarizer, which operates in the medical domain, a domain particularly well-suited to the problem of providing personalized information. Evaluation with subjects shows that user-sensitive summaries help access relevant information more efficiently and accurately than generic summaries of the same material.

The Colloquium is supported by generous contributions from the Bloomberg, Information Builders, Inc. and Netlogic, Inc.