BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Cross-lingual Learning\, and Applications in Dialog and Translatio
 n - Junjie Hu\, University of Wisconsin-Madison
DTSTART:20220609T140000Z
DTEND:20220609T150000Z
UID:TALK175475@talks.cam.ac.uk
CONTACT:Marinela Parovic
DESCRIPTION:Over the last decade\, the phenomenal success of NLP systems h
 as been mostly driven by deep neural networks and supervised machine learn
 ing approaches to a large amount of labeled data. However\, it’s infeasi
 ble to annotate data under all possible real-world scenarios. As a result\
 , these systems may fail dramatically in practice when dealing with comple
 x textual data written in different languages\, or even associated with di
 fferent data modalities. In this talk\, I will present works that are impo
 rtant to extend the generalization ability of NLP systems across languages
 . First\, I will present our work on XTREME which provides a platform for 
 cross-lingual learning on 9 NLP tasks in 40 languages. ​​I will then d
 emonstrate two cross-lingual applications of task-oriented dialog and mach
 ine translation. This talk will be concluded with an overview of my resear
 ch and some future directions.
LOCATION:https://cam-ac-uk.zoom.us/j/97599459216?pwd=QTRsOWZCOXRTREVnbTJBd
 XVpOXFvdz09
END:VEVENT
END:VCALENDAR