An ARC-funded PhD scholarship, tenable at either the University of
Sydney (Linguistics Dept) or the Flinders University of South Australia
(Computer Science Dept), is available for research on exploring the
application of unsupervised learning techniques to Japanese. Applicants
should have either a good Honours degree or an equivalent four year
undergraduate degree in an appropriate field, and be seeking to
undertake full-time research leading towards a PhD. The project is in
the area of using machine learning and statistical methods to
inductively acquire knowledge of the structure of Japanese (in line with
the grammar discovery approach of structural and corpus linguistics),
and then to be able to apply that knowledge to practical problems. This
is a collaborative project and opportunities exist to spend time at both
institutions. Both have active research programs and up-to-date
facilities for work in this area. Required qualifications are (i) good
programming ability, (ii) familiarity with Chinese character-based
writing systems, and (iii) an interest in doing research in this area.
Desirable qualifications include: (i) background in machine learning,
(ii) background in linguistics/NLP, (iii) knowledge of corpus-based and
statistical learning approaches, and (iv) knowledge of Japanese.
Applicants with research interests and background across these areas are
particularly encouraged to apply. The scholarship is valued at the
standard APA (1998) rate of $15,888 p.a. (tax exempt) for 3 years.
Further details can be obtained from Christopher Manning (Dept of
Linguistics, University of Sydney, +61 2 9351 7516,
or David Powers (Dept of Computer Science, Flinders University,
+61 8 8201 3663, firstname.lastname@example.org,
including a full curriculum vitae and the names and contact info (post,
email, fax) of at least two referees, should be sent either by email to
the above addresses or to Christopher Manning, Linguistics F12,
University of Sydney NSW 2006, AUSTRALIA by the closing date.
Closing: Direct, 29 May 1998 (or until filled)