|
Courses and Workshops
|
Description The course will be structured as follows:
* linguistic aspects (mainly transfer problems, ambiguities, lexical
gaps etc.)
* computational problems of machine translation
* paradigma of machine translation (direct, transfer, interingua)
* statistical based MT
* example-based MT
* evaluation of Machine translation
* examples of MT Systems
The course is addressed to students that have minimal knowledge of
linguistics, computational linguistics, and some programming knowledge
(Prolog and or Java desirable) The second part of the course will be
applicative, the students will be divided in groups and asked (and
guided) to implement a small demonstrative MT-system. In this way the
students will be aware with the challenges of the implementation of an
MT-System.
/Further particulars/
Cristina Vertan received her Ph.D in Computer Science in 2000, from
University of Bucharest. She has also a Master degree in Statistics from
the Free University in Brussels In 2000 she worked as researcher in the
project Verbmobil-speech to speech translation and developed the last
version of the evaluation tool. Between 2001 and 2002 she was a Humboldt
Fellow at University of Hamburg. Actually she has a senior researcher
position at the same University. Research and teaching activites:
machine translation, evaluation of machine translation, architectures
for language processing systems. The proposed course is based on
severeal semester teaching experience at University of Hamburg
|
© 2003, 2004 LORIA, Carlos Areces. |
2003-11-26
| |
