THE INNOVATIVE
GOALS OF THE PROJECT:
- to develop a generic and implementable model
of natural language interaction
- to test and compare various soft computing
methods for interaction modelling and for development of a working system
- to develop a working system which can be intergrated
into various applications that require natural language interaction
One of the important aspects of the project is
to strengthen Finnish technology in natural language processing, building
upon such internationally acclaimed techniques as Kimmo Koskenniemi's
two-level model for morphological analysis and Teuvo Kohonen's self-organising
maps for clustering data into meaning information classes.
The rapid development in IT has caused computer
interaction become more common, so the need for flexible interaction
is apparent not only in everyday computer use but also when the user may
not be familiar with the situation. The system should thus be able to
adapt to various situations and various users, so as to provide most efficient
and helpful mode of interaction. This also implies that the special needs
of disabled people will be taken into account when designing more natural
interactive systems for access to information.
The explosion of available information also requires
that the systems should be able to deal with the problem of knowledge
acquisition, and in this respect systems that can learn are of great
interest. The capacity of computers has also increased, enabling rapid
testing and comparision of various machine learning techniques. The project
uses various soft-computing and learning techniques to automate and build
a learning system that will also learn from its interaction with the
user.
The achievements of the project were presented in a final demo event on
12/12/2002.
Contact Dr. Kristiina Jokinen,
kjokinen@uiah.fi for more information.
<< Back
to the summary
|