Date: June 05, 2003 Time: (All day)
Event Type: Lecture
In recent years, NLP research has focused primarily on the development of techniques for improved multilingual processing. The conclusion drawn by a growing number of multilingual NLP researchers is that progress in the field requires access to both statistical and linguistic knowledge. I will illustrate the realization of this conclusion by describing research and development on two specific multilingual NLP tasks: (1) Divergence Unraveling for Machine Translation; and (2) Headline Generation for translingual retrieval. I will introduce some examples of systems developed for these tasks and then present some experimental results.