Natural Language Processing
University of Melbourne, 10-14 July 2006
|
|
Natural Language Processing University of Melbourne, 10-14 July 2006 |
|
This tutorial will begin with a thorough introduction to the FrameNet lexical database, which provides very detailed information about the syntactic and semantic combinatorics of roughly 10,000 English words, on the basis more than 100,000 sentences hand-annotated along Frame Semantic principles. We will delve into the inner structure of the data including frame-to-frame and frame element-to-frame element relations (both across frames and within frames) and semantic typing on frame elements, and discuss links to ontologies and other lexical resources. We will then explore both current and prospective uses of FrameNet in NLP tasks (such as text understanding and paraphrase generation), and also the application of NLP techniques to partially automate the development of the lexicon itself. Students will gain a better understanding of how to use lexical databases in general and FrameNet in particular in building an NLP system.
Collin F. Baker (PhD. Linguistics, UC Berkeley 1999) has worked with the FrameNet project at the International Computer Science Institute since 1997, and as Project Manager since 2000, with funding from the National Science Foundation, DARPA, and ARDA. His current research focuses on the evaluation of systems for computer-assisted frame discovery and automatic semantic role labeling. He is a member of the ACL and the LSA and has taught English as a Second Language in Taiwan, Japan, and the United States.