[Elsnet-list] Studentship - "Natural Language Generation for dialogue systems"

Gillian Callaghan g.callaghan at dcs.shef.ac.uk
Thu Apr 29 09:42:38 CEST 2004

Please advertise the following studentship.  Apologies for multiple

PhD Studentship - "Natural Language Generation for dialogue systems"

The Department of Computer Science at the University of Sheffield invites
applications for a three-year studentship award to commence in October 2004.
The successful applicant will work on Natural Language Generation for
dialogue systems and will be jointly supervised by Prof. Marilyn Walker and
Dr. Mirella Lapata. 

Deadline for applications: Thursday 20 May 2004. 

Bootstrapping Natural Language Generation in Dialogue Systems

In the past few years there has been an increasing interest in the use of
natural language generation in spoken dialogue. The spoken language
generator is a component of the dialogue system that, given a high level
goal of what the system is supposed to say, decides what the system should
say to achieve the goal (sentence planning) and how to say it (sentence

Most current dialogue systems rely on hand-crafted rules or templates for
performing the generation task, therefore requiring considerable manual
effort to produce meaningful output. These systems tend to be domain- and
situation-specific and typically have to be entirely re-engineered when
switching to a new application domain. 
The project will investigate data intensive techniques for automating the
task of sentence planning and realization for dialogue systems while
avoiding the need for manual hand coding. The general aim is to develop
automatic methods for creating natural language generation components that
produce quality output efficiently and can be easily adapted to different
domains and dialogue situations. 

Applicants should have a good honours degree or equivalent in Computer
Science, Computational Linguistics or Linguistics. Applicants with strong
programming skills (e.g., C++, Perl, Java) and expertise in any of the
following will be preferred: 
. Dialogue modelling 
. Statistical modelling 
. Natural Language Processing 
. Machine Learning 
. Grammar formalisms (in particular, dependency-based ones) 
. Theories of discourse structure 
. Pragmatics (in particular, speech act theory and implicatures) 

The successful applicant will be based at the University of Sheffield's
Department of Computer Science, internationally recognised for the quality
of its research, and will work across the Natural Language Processing (NLP),
Speech and Hearing (SpandH), and the Machine Learning (ML) research groups
(see http://www.dcs.shef.ac.uk/research for details). 

The Department of Computer Science has close research links with a number of
other academic institutions (e.g., University of Edinburgh, MIT) and
companies (e.g., AT&T) from which the student will benefit. For more
information about the Department see http://www.dcs.shef.ac.uk/. 

Information about the PhD Programme, and how to apply for a PhD can be found
by following the various links from the following URL:

PLEASE MARK "Bootstrapping Natural Language Generation in Dialogue Systems"
ON THE APPLICATION. For additional information on the PhD project, please
contact Prof. Marilyn Walker (m.a.walker at sheffield.ac.uk) or Dr. Mirella
Lapata (m.lapata at dcs.shef.ac.uk). 

An application form, and details on the method of application, are at

More information about the Elsnet-list mailing list