[Ltg] LTG Seminar [James Curran, Nov. 15, 11am, E6A]

[Rolf Schwitter]

----
LTG Seminar
 - see: http://www.clt.mq.edu.au/Events/Seminars.html

Monday, November 15, 2004 at 11am
Macquarie Uni, E6A 357
----

Speaker: James Curran
Title: Maximum Entropy models for Natural Language Processing

Maximum Entropy models, in their various forms, have been successfully 
applied to a wide range of Natural Language Processing problems including 
tagging, parsing, and classification tasks.

In this talk, I will avoid describing the sophisticated theoretical 
foundations of Maximum Entropy modelling in favour of presenting a clear 
intuitive understanding.  Consider this a taster for my ALTA Summer School 
course which will cover both the theory and intuition in detail.

I will then describe a range of applications that Stephen Clark and I have 
obtained state-of-the-art results for using MaxEnt models, including:
  * POS tagging
  * Named Entity Recognition
  * Supertagging for Combinatory Categorial Grammar

I will also talk about some ongoing experiments that my students and I have 
just completed using MaxEnt modelling, including:
  * learning about unknown words using large raw text corpora
  * question classification for QA systems

Hopefully, I will have time to share some of the practical knowledge that we 
have discovered along the way which will help you apply MaxEnt modelling to 
your own problems.

Finally, I will talk about some directions and applications where I see 
MaxEnt modelling going next.


Bio:

James Curran is an ARC Postdoctoral Fellow in the Language Technology 
Research Group in the School of Information Technologies at the University 
of Sydney. He has just returned to Australia after completing his Ph.D. in 
computational lexical semantics at the University of Edinburgh.

His ARC funded project, 'Ask the Net: Intelligent Natural Language 
Learning', involves automatically asking contributors simple questions via 
email which will be collected to create annotated data for standard NLP 
problems, e.g. Named Entity Recognition.  An interesting challenge is 
finding ways of eliciting linguistic knowledge from those without lingustic 
training.  His other research interests range from standard statistical NLP 
problems such as tagging and parsing, through to system building such as the 
development of question answering and biological text mining systems.