perm filename KOL[AM,DBL] blob
sn#568259 filedate 1981-03-05 generic text, type C, neo UTF8
COMMENT ā VALID 00002 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00002 00002 (a) Professor Janet Kolodner's research has already demonstrated the deep
C00007 ENDMK
Cā;
(a) Professor Janet Kolodner's research has already demonstrated the deep
connections that exist between the way information is organized in memory
and the modes in which its retrievals are and are not efficient. She has
carried this out not with a purely theoretical model, but rather with an
operational one, a computer model which can be used as an experimental
vehicle on which to perform experiements. Her results to date should have
an impact on data base organization and searching, on aritificial
intelligence, and on cognitive psychology.
(b) Dr. Kolodner emerged from the Yale University Cognitive Science group,
a group strongly dominated by one particular research programme. To her
credit, she independently discovered, developed, and used a hierarchical
frame-like representation scheme in her research, and convinced the other
members of that group of its utility. This scheme -- which is now
referred to as MOPS at Yale -- enabled synergy between the "scripts" and
"frames" methodologies. As a second case, let me remark upon the work
Kolodner performed as a summer student two years ago, at IBM Yorktown
heights, New York. She was alone in an environment which was largely
ignorant of Artificial Intelligence and Cognitive Science research to
date, and skeptical or outright hostile to what parts of it they had been
exposed to. Kolodner managed to convince the key personnel of the
importance of trying to understand what the user of their system meant,
rather than focusing on superficial but more concentional approaches. At
the end of the summer, the head of that project, Dr. Maurice Karnaugh (of
Karnaugh maps) took a semester off from IBM, to spend studying AI at Yale.
He was made the head of a tiny -- but novel -- AI group at IBM. In
summary, I believe Kolodner to be capable of extremely independent,
creative acitivities, and her research to date leads us to expect great
things from her in the coming years. This brings us to:
(c) The intended research is precisely the right thing for her to carry
out at the present time: experimentation upon the system she built as her
doctoral dissertation. By experimenting with it, she will hopefully be
led to the next great breakthrough in memory organization. Kolodner has
avoided the temptation to weave a theory out of midair and then fumble
about trying to verify it; rather, she is willing to exploit the
experimental vehicle she has -- and others to come -- and be guided
dynamically by the empirical data that results. This is the hallmark of
great science, and I am waiting eagerly to hear how it progresses.