Report on UIAH Media Lab's doctoral seminar
- place: Media Lab, lecture room
- date: 20 Jan, 2000
Natural language interaction
Timo informed that the project on natural language interaction
(http://www.mlab.uiah.fi/vuorovaikutus/) has a reasonably good
chance to become funded in Tekes' USIX programme. A full-time
research position will be available for a person having
expertise in the area of natural language processing but
also part-time works related to interaction design of such
systems would also be available. For instance, Antti would
prepare a report from the point of view of the Deaf.
Next NIMRES
Next Nordic Multimedia Research School will be held
in Gothenburg (Göteborg) on June 2-9 with the theme
"Real and Virtual Communities". Themes incluce
"communities in urban context", "virtual communities",
"communities of virtual agents" and "simulations and
virtual environments for children".
In-depth questions: relevance, rationality
A long conversation was going through many kinds of in
depth questions related to the initial topics of (natural
language) interaction and relevance. Jukka pointed out
the Socratic method and the question if this could be used
in natural language processing. (See, for instance:
http://sun.soci.niu.edu/~phildept/Dye/method.html)
Timo pointed out that conducting good conversation requires
substancial amount of background knowledge. This has also
been pointed out by by Joseph Beck, Mia Stern and Erik Haugsjaa
in their article on "Applications of AI in Education":
"Another obstacle in representing multiple teaching strategies is the
limitations imposed by other components of the ITS [intelligent
tutoring systems] in addition to those placed by the student model.
In particular, the difficulty of representing knowledge impedes the
ability to explicitly represent teaching strategies. For example, the
Socratic method requires substantial ``common sense'' knowledge beyond
what is maintained in a domain knowledge base. This problem of
scaling up the knowledge base is not unique to ITSs and is common to
many areas of AI." (http://www.acm.org/crossroads/xrds3-1/aied.html)
We discussed about issues related to the world views. Lily pointed
out some problems related to the concept of rationality. Jukka
mentioned that earlier in the philosophy it was thought that
science has replaced myth in its rationality but nowadays the
view is not that straightforward. Lily referred to the problems
of categorisations and related to that Timo mentioned how rule-based
systems lead into straightforward associations even though only
a slight correlation might exist. Conception of rationality has
traditionally been bound to words and logic which may lead to
even to various forms of racism (the text at the page
http://www.mlab.uiah.fi/~timo/runo1.html
may illuminate the link).
Next times and place
We decided that the next times will be held at 16 and the lecture
room will be abandoned because it tends to cause headache for several
people. It was discussed if the reason is in the lights or in the
whistling background noise (or something else). The next time will
be held either in the meeting room of ML or in Timo's room.
There will be short presentations of the research topics
(conversations on general themes will also be continued):
- 3.2.: Kari Jääskeläinen
- 17.2.: Minna Väisänen
- 2.3.: Lily Diaz
Timo Honkela,
20 Jan 2000