SALS-SIG Research Seminar

NLP Research at the Microsoft Institute

Robert Dale
Microsoft Institute

When: Tuesday, 17th May 1994

Time: 4:00pm

Where: Room E6A357, Macquarie University

Abstract:

The Microsoft Institute of Advanced Software Technology, in a Partnership for Development program with the Australian Government, aims to foster the Australian software industry through technology transfer and developer assistance. One element of this strategy provides for basic research to be carried out at the Institute, and so the Institute has put in place two research groups working on Natural Language Processing and Distributed Computing.

In this talk, I'll give an overview of the research in Natural Language Processing at the Microsoft Institute. Our work here falls into two main themes: natural language generation, and the use of large on-line textual resources for natural language processing tasks. I'll describe the work going on in each of these areas:

• in generation, we are looking at the generation of referring expressions, and the implications of some of this work
• for architectural issues in the construction of NLG systems; and
• in the on-line resources area, we are looking at performing various kinds of disambiguation---prepositional phrase attachment, compound noun structure, and sense disambiguation---using information derived from resources such as corpora and machine readable dictionaries.

Enquiries: Maria Milosavljevic 9850 6345 mariam@mpce.mq.edu.au