The Department of
Computer Science at the
Improve upon our national rating as a research department. Our department has been consistently rated 11-13 by surveys in U.S. News and World Report and the National Academy of Sciences. The faculty is very strong; every junior faculty member we have hired in the past fifteen years has won an NSF CAREER award. Many senior faculty are members or founding members of major technical societies (IEEE, ACM AAI, etc.) and have been honored for their achievements in many ways. Funding from all sources‑especially NSF‑is at all time highs. We have also been exceptionally successful with our submissions to NSF’s ITR program. Our primary goal is to move the Department into the top 10 nationally. But the competition is fierce. Challenges:
·
Place our
best graduate students in
better academic positions. Our record here has been improving, with recent
graduates currently in tenured and tenure track positions at Brown, Rutgers and
Toronto.
·
Identify resources,
especially, adequate laboratory and office space, that will enable the
Department to move into strategically important areas of research in a very
competitive national environment.
·
Improve the
services offered by our graduate office, which is currently significantly
understaffed and operates using many of the same processes that were in place
in the early 1980’s.
·
Improve our collaboration with the architecture,
systems and robotics faculty in the Electrical
and Computer Engineering Department to develop strong research programs in
embedded computing and robotics.
Enrich our
undergraduate education program
by expanding the research opportunities for talented undergraduates, expanding
and improving our project-oriented courses and making more extensive use of
local industry and government laboratories in delivering undergraduate
education. The Department is embarking on a major renovation of its
undergraduate education program, redesigning many of the courses and bringing
tenure track faculty into the classroom for all of the introductory courses. Challenges:
·
“Customer
support” for undergraduates.
Our undergraduate office is understaffed, with inadequate manpower for
advising, tutoring and working with students on extracurricular activities
·
Space for
undergraduate activities.
The Department has no undergraduate lounge, no space to develop new
laboratories for undergraduate education and research, and very limited space
for teaching assistants to meet with students and help them with projects and
technical questions. For our 1700 undergraduate students we have less than 300
square feet assigned to this function, and it is not uncommon to find students
camped out in hallways waiting for their (brief) turn for assistance.
Develop far
better relations with our “communities.” Our Department has not done a good job of maintaining ties with its
alumni or with local and national industry (although many individual faculty do
very well in establishing research relationships with industry). It has not had
an active advisory board in many years. This is in large measure due to severe
staff shortages in the chair’s office, aggravated in the past year by budget
rescissions that have resulted in losing two important staff positions. Challenges:
·
Address the staff
shortage by adding at least two new staff positions to the chair’s office –
one for public relations, alumni relations and fund raising, and a second as an
administrative assistant to the chair.
·
Reconstitute
the Department’s Advisory Board
as a first step towards developing new ties to industry and government. We must
make better use of both senior management in local industry and government, as
well as some of our more successful alumni in industry, academia and
government.
·
Identify and track our most successful alumni – both as a source
of potential gifts to support department programs, and to document the success
of our graduates at all levels.
The remaining
sections of this report on Research, Undergraduate Education and Graduate
Education discuss these challenges at greater length in the context of the
Department’s overall activities and strengths.
It has been over ten years since the Department
was last reviewed. Both the internal and external reviews are available at Past Reviews. Here, we list the most critical
weaknesses identified by the last external review committee, and provide our
own assessment of whether these issues have been adequately addressed during
the intervening decade.
1. Lack of
consensus and ineffective governing structure. This was a serious problem
at the time of the last review, which we believe has been largely solved,
although without having to make any significant changes to the governing
structure of the Department. In part, the solution was forced upon us several
years ago when the campus required each Department to construct five-year
strategic plans that identified the most important areas for growth, and plans
for moving into those areas. In conjunction with UMIACS, the Department
initially identified thrust areas of computer systems and software engineering
and programming languages and recruited vigorously and successfully in these
areas. While current resources for recruiting are severely limited by the
overall financial condition of the State University system, there is consensus
that we should build a computationally biology group (and associated education
programs at both the graduate and undergraduate level). The review committee
also identified the structure of the graduate program as another example of
failure to achieve consensus (every faculty member had some complaint about it,
but the structure persisted). Since that time we have completely re-organized
the graduate program, and are in the process of doing so again this year.
2. Excessive dependence on counting in
evaluations, as opposed to “impact.” The external review identified
reluctance on the part of the faculty to evaluate research on the basis of
long-term impact. We do not believe that this is the case today. The
Department's promotion process places especially high emphasis on external
letters of recommendation, and we strongly urge our references to comment on
the impact of the research conducted by our faculty. The importance of
publications in highly refereed conferences is appreciated both at the
Department and higher levels of our campus during the promotion process.
3. Slow to move
into new areas. Prior to the 1992 review, and for several years afterwards,
the Department was essentially unwilling to offer positions to senior
candidates, clearly the best way to move into new areas. This was mostly a
reluctance to stray from the strategy that had worked in the past—hiring
talented junior people and letting them “grow” on the job. The situation is now
completely reversed. We did hire Joel Saltz as an Associate Professor shortly
after the review (and this did allow us to move quickly and effectively into
the high performance computing area. This is an area in which the Department is
now strong, so that we have been able to maintain our strength even with Joel’s
recent departure for Ohio State), and we have also made several senior offers
to candidates in graphics, theory, scientific computing and computer vision.
Within the past three years we have hired a “senior” Assistant Professor with a
short time frame for consideration for tenure (Iftode), one untenured Associate
Professor (Varshney‑who was quickly tenured), two tenured Associate
Professors (Srinivasan and Jacobs), and one Full Professor (DeFloriani).
4. Dependence
on UMIACS. The external review committee felt that the Department was not
able to control its budget because of its dependence on UMIACS. The current
chair was then UMIACS director. There were sources of tension between the Department
and the Institute, mostly related to issues of permanent appointments and
teaching loads for Computer Science faculty who did not have appointments in
research institutes. Moreover, UMIACS was a relatively new entity at that time
and was given more flexibility in budgetary decisions than the Department. These
problems no longer exist. During the past few years the Department has received
significant enhancement funds for hiring new faculty in systems, graphics, HCI,
and software engineering. It has received major new funding to support its undergraduate
education program several times over the past 6–7 years. It (most importantly
our previous chair, Prof. Gannon) has lobbied successfully for the construction
of the new classroom facility for computer science education. A good amount of
this funding has been lost during the current budget downturn, but the
Department is in a far better position financially than it would have been
without the additional support it received from the campus during the past five
years.
5. Graduate
program not a high priority. We have made significant changes (which we
believe are improvements) to our graduate program. It is an area in which we
need to make much more progress. The old written comprehensive exam system has
been replaced by a course-based system. This, of course, solved some problems
and introduced a whole new set. The program is under review once again based on
discussions at our Department retreat in January 2001. But many of the
deficiencies noted in the last review remain - mentoring of new graduate
students (although we have instituted faculty/student lunches, seminars on
academic careers, and other mechanisms to increase contact between new grad
students and faculty), actively recruiting new graduate students, and, most
important, placing our Ph. D.'s, are all areas in which significant improvement
is still necessary.
6. Joint
appointments. The last review suggested that additional joint appointments
with ECE would be beneficial. This was symptomatic of a dearth of
interdisciplinary research and educational activities involving computer
science faculty. There has been astounding progress in this area over the past
decade. Our faculty is now involved in an amazing variety of interdisciplinary
research projects with faculty from Electrical and Computer Engineering,
Nuclear Engineering, Linguistics, Sociology, Library and Information Sciences,
Mathematics, Physics, Business and Management, and Geography. The table below
lists our affiliate faculty, showing their home departments and the faculty in
our department whom they work with. There are many additional relationships
involving UMIACS faculty and others without appointments in either Computer
Science or UMIACS.
Affiliate
|
Home
Department
|
Collaborates
with
|
|
John Baras |
ECE |
Roussopoulos |
|
Rajee Barua |
ECE |
|
|
Shuyra
Bhatacharya |
ECE |
|
|
|
ECE |
Davis,
Aloimonos, Jacobs |
|
Michael Cukier |
Materials and
Nuclear Engineering |
Arbaugh, Basili,
Zelkowitz |
|
Allison Druin |
Library and
Information Studies |
Bederson,
Shneiderman, Gumbrietiere |
|
Manoj Franklin |
ECE |
|
|
John Grant |
Towson State
Math |
Perlis |
|
Satyandra Gupta |
Mechanical
Engineering |
Nau |
|
John Horty |
Philosophy |
Perlis |
|
|
UMIACS/ECE |
Roussopoulos,
Sussman |
|
Gang Qu |
ECE |
|
|
Louiqa Raschid |
Business and
Management |
Roussopoulos,
Dorr |
|
Phil Resnick |
Linguistics |
Dorr |
|
Uzi Vishkin |
ECE |
Tseng |
|
Weinberg |
Linguistics |
Dorr |
|
Donald Yueng |
ECE |
Tseng |
Additionally,
progress is being made on interdisciplinary undergraduate education programs. We
recently completed an agreement with Physics for an undergraduate track in
computational physics and are in the midst of the design of an undergraduate
joint major in computer science and biology.
7. Minorities.
The external review felt that the Department was not sufficiently active in
recruiting minority students‑both undergraduates and graduates. The
University as a whole is now very successful at recruiting undergraduate
minorities, and the Department still does not play any role in this process. Unfortunately,
no steps have been taken to increase recruiting of minority graduate students.
In November 2001, 20 women and minority graduate students from the US
and Canada came to our campus for a workshop entitled “Research, Careers, and
Computer Science: A Maryland Symposium.” The participants presented their
research, met Maryland faculty and students, and participated in interactive
sessions about academic jobs. These sessions focused on teaching, research,
obtaining funding, the tenure process, distinctions between academic and other
jobs, and time management. This event was sponsored by both the Computer
Science Department and UMIACS. More information can be found on the workshop
website: http://www/cs.umd.edu/users/oleary/workshop/.
8. Salary compression. Our Department
still suffers from severe salary compression, with some assistant professors
having higher salaries than the lowest paid full professors. Starting salaries
and startup packages are nationally competitive, but our senior faculty have
fallen further behind since the last review. This is summarized in the table
below that contains a Taulbee-like survey of Maryland salaries versus national
statistics.
|
|
Assistant
Professor |
Associate
Professors |
Full Professor |
||||||
|
|
Min |
Mean |
Max |
Min |
Mean |
Max |
Min |
Mean |
Max |
|
Maryland |
79,500 |
84,402 |
93,600 |
67,915 |
94,194 |
105,300 |
92,427 |
120,524 |
172,869 |
|
National (1-12)
means |
74,711 |
80,891 |
86,483 |
84,148 |
91,412 |
97,949 |
88,632 |
122,732 |
168,860 |
|
(13-24) |
78,070 |
83,673 |
90,538 |
85,663 |
92,985 |
98,827 |
94,322 |
127,845 |
185,306 |