(approximately $500). The Council
also discussed the creation of a named
award with an endowment to support
future awards. In addition, they recommended the award be listed by
department and not just by university.
Dietterich noted that the approval by
ACM will take a long time, and that
they hope to start that process in
March or April so that nominations
could be submitted for 2018. He noted
that this award would be a subgroup
award, so would not have the same
prestige as the ACM Doctoral Dissertation Award. There was a brief discussion about having an award sponsored
by AAAI only, but the group was supportive of the joint venture and noted
that it was in line with similar efforts,
such as the Doctoral Consortium.
Conference Committee Report
Shlomo Zilberstein reported that he
has program chairs in place for the
next two years. Sheila McIlraith and
Kilian Weinberger will serve in 2018,
and Pascal Van Hentenryck and Zhi-Hua Zhou in 2019. Future chairs will
have the opportunity to observe in
2017 and 2018. He also noted that the
program for the current 2017 conference was very rich and exciting. Technical submissions, including main and
special track submissions, rose by over
20% to 2,595, excluding senior member papers, demos, doctoral consortium, and student poster abstracts.
Thirty percent of submissions came
from China, with acceptance rates for
these papers and US-based papers sim-
ilar to overall acceptance rates. Ma-
chine learning and natural language
processing, were the leading subject
areas, with strong representation from
applications, search, vision, knowledge
representation and reasoning, game
theory and economic paradigms, and
multiagent systems. The Council dis-
cussed the issue of authors submitting
multiple papers, and whether a cap
should be instituted. In some cases,
this practice has become quite signifi-
cant, with one author submitting more
than 40 papers in 2017. The Confer-
ence Committee expressed concerns
about load balancing and fairness to all
authors. There has been some backlash
from the community when strict limits
have been imposed. Shlomo noted
that the double-blind review helps mit-
igate the problem of multiple submis-
sions, so should be continued. The
overall consensus was that no formal
policy will be instituted, but program
chairs may wish to examine things
more closely at the time that the CFP is
written.
Zilberstein also examined the changing character of the conference with its
significant growth in recent years. He
solicited ideas for ways to address this
change. Several suggestions were made,
including organizing “birds of a feather” meetings, expanding the number of
parallel sessions, expanding the focus
of the conference beyond the traditional academic one, expanding our
outreach efforts, holding sessions
devoted to policy issues, and increasing
opportunities for socializing. The group
discussed the current scheduling app
and web program, Guidebook, and
agreed that it lacked good search tools.
Blai Bonet agreed to help determine if
its usability could be improved for
2018, as this will become even more
important as the conference continues
to grow. The Council generally agreed
that they want to retain the current
model, but explore ways to change and
grow with coexisting new programs,
such as AI in Practice.
Zilberstein noted that the AI in History program will continue in 2017
with a panel chaired by Ed Feigenbaum
on Expert Systems. The original plan
was to have something every two
years, starting with the Shakey celebration in 2015. In order to sustain this
activity, Shlomo suggested that a committee be formed to select the next
focus. The Council suggested that the
program chairs for 2019 put a subcommittee together to address this issue.
Zilberstein discussed the current
travel restrictions put in place by the
US government, and the resulting
uncertainty for the future. Shlomo
encouraged the Council to adopt a policy that would help affected researchers, such as allowing remote
attendance, or in-person presentation
at a related conference such as IJCAI.
While the Council is generally in favor
of finding reciprocal solutions, they
agreed that further study and evaluation was required to make an informed
decision about future procedures.
Finally, the Council discussed the
ongoing issue of finding a good algorithm for assigning papers, especially
in light of the dramatically increasing
number of submissions and the growing program committee. Zilberstein
noted that Shaul Markovitch had created software to help with the assignment algorithm. In conjunction with
this, the group discussed whether it
made sense to change submission systems. Zilberstein suggested that another comprehensive conference survey
might be useful.
Finance
Ted Senator announced that David
Smith will be taking over for him as
Secretary-Treasurer in early 2018. They
have entered a transition period
whereby Smith will be involved in all
financial discussions to allow him time
to become familiar with AAAI’s
finances and financial policies. Senator
reviewed the general financial principles of AAAI, including its policy of
spending up to 3-5 percent of the operating reserve on an annual basis. He
noted that the investments have continued to grow over time, and that
AAAI is now enjoying a surplus after
several years of effort to reduce an
annual budget deficit. There are specific targets in place for the various
funds in the investment portfolio, and
the general practice is to not shift
funds frequently but to maintain targets by withdrawing from funds that
have become out of line. In response to
concerns about the annual circulation
of the tax return to members of the
Council, Senator explained that the
main purpose of this exercise is to
maintain transparency. It is understood that Council members cannot be
expected to verify all the information.
International
Qiang Yang asked the Council for
input on effective ways to promote AI
worldwide. While there has been some
success in China and India, European
efforts have been less organized, and AI
work in that area has not been well
promoted. The Council agreed that the
make-up of the committee should be
as international as possible. Several
suggestions were made, including
jointly sponsoring meetings in Europe,
WINTER 2017 109