Time: 1245pm-145pm, Friday, Feb 19
Place: Room 6496, CUNY Graduate Center, 365 Fifth Ave (34str&35str).
Speaker: Zheng Chen (CUNY)
Title: Can One Language Bootstrap the Other: A Case Study on Event Extraction

Abstract: 

We propose a new bootstrapping framework using cross-lingual information 
projection. We demonstrate that this framework is particularly effective for a 
challenging NLP task which is situated at the end of a pipeline and thus suffers 
from the errors propagated from up-stream processing and has low performance 
baseline. Using Chinese event extraction as a case study and bitexts as a new 
source of information, we present three bootstrapping techniques. We first 
conclude that the standard mono-lingual bootstrapping approach is not so 
effective. Then we exploit a second approach that potentially benefits from the 
extra information captured by an English event extraction system and projected 
into Chinese. Such a cross-lingual scheme produces significant performance gain. 
Finally we show that the combination of mono-lingual and cross-lingual 
information in bootstrapping can further enhance the performance. Ultimately 
this new framework obtained 10.1% relative improvement in trigger labeling 
(F-measure) and 9.5% relative improvement in argument-labeling.

Speaker Bio: 

Zheng Chen is a Ph.D. student in Computer Science at the Graduate Center, the 
City University of New York. He is a member of BLENDER lab directed by Prof. 
Heng Ji. His research interests generally lie in computational linguistics and 
statistical machine learning, especially, cross-document cross-lingual 
information extraction, i.e., how to identify important facts (entities, 
relations, events) from web-scale corpus, how to resolve multiple mentions of 
the same entity/event. He has published 6 papers at NLP conferences, such as 
NAACL, ACL.