[ont.events] UW AI Seminar, Prof. Cercone on "The Automated Academic Advisor: An Introduction and Initial Assessment"

mwang@watmath.UUCP (mwang) (05/08/84)

_D_E_P_A_R_T_M_E_N_T _O_F _C_O_M_P_U_T_E_R _S_C_I_E_N_C_E
_U_N_I_V_E_R_S_I_T_Y _O_F _W_A_T_E_R_L_O_O
_S_E_M_I_N_A_R _A_C_T_I_V_I_T_I_E_S

_A_R_T_I_F_I_C_I_A_L _I_N_T_E_L_L_I_G_E_N_C_E _S_E_M_I_N_A_R
                           - Friday, May 18, 1984.

Prof. N. Cercone of Simon Fraser University will  speak
on  ``The  Automated  Academic Advisor: An Introduction
and Initial Assessment.''

TIME:                11:30 AM  (Please Note)

ROOM:              MC 2035

ABSTRACT

It is increasingly difficult for academic  advisors  to
maintain  current  tracking records on each student and
to stay informed about  graduation  requirements  which
frequently change.  A large practical artificial intel-
ligence system which accepts natural language  queries,
the  Automated Academic Advisor (AAA), is currently be-
ing developed in our Laboratory for Computer  and  Com-
munications Research (LCCR).  The AAA will offer facul-
ty, students, and academic advisors  accurate  informa-
tion  on degree requirements and course offerings.  Two
parsers are described which were designed for  ue  with
the  AAA.   The  ATN parser ENGRA is described with em-
phasis given to several enhancements incorporated  into
ENGRA.   Our first AAA prototype incorporates ENGRA to-
gether with the semantic interpreter (SI)  to  generate
formal SQL queries for the ORACLE relational data base.
SHADOW, a Prolog-based English analyzer which forms the
basis  of our second prototype, is then described.  In-
formal comparisons are made between  the  applicability
of  the  two parsers to the AAA and our initial experi-
ence with these two prototypes is discussed.   The  AAA
is  unique  in  that the impact and performance of this
interactive information system will be evaluated in  an
organisational  setting  - our student advisory system.
The AAA will be actively involved in its own evaluation
(cf.  Telidon).  The design of the evaluation subsystem
is discussed briefly with the emphasis  on  `universal'
techniques  of system evaluation which will permit con-
sistency and comparability of evaluation.  Our  evalua-
tion emphasis is placed on the system's quality assess-

                      May 8, 1984

                         - 2 -

ment rather than the more traditional performance meas-
urement criterion.

R. Hadley, F. Martin, and T. Strzalkowski have  contri-
buted  significantly  to  the  ideas  presented in this
talk.

                      May 8, 1984