rick@cs.arizona.edu (Rick Schlichting) (02/03/91)
--------------------------------------------------------------------------
Trip Report -- Shinshu University (July 2, 1990)
Richard D. Schlichting (rick@cs.arizona.edu)
Associate Professor
Department of Computer Science
The University of Arizona
Tucson, AZ 86721, USA
[This report represents the personal opinion of the author, who was on
sabbatical in Japan from Dec. 1989 through July 1990. The sabbatical
was supported in part by grant INT-8910818 from the NSF U.S.-Japan
Cooperative Science Program.]
---------------------------------------------------------------------------
On July 2, I visited the Faculty of Engineering campus of Shinshu
University in Nagano City. My host for the visit was Prof. Masayuki
Okamoto of the Department of Information Engineering. Prof. Okamoto
is a former student of my host scientist at Tokyo Institute of Technology,
Prof. Takuya Katayama, and had been at the department at Shinshu since its
founding in 1978 or so. After arriving at the university, he first gave
me a brief rundown of the department. There are currently 7 chairs in
the department, which is organized like most Japanese universities into the
chair system (koza in Japanese.) In his lab, there is currently one
associate professor (working in theory of computing), and about 15 MS
students. There is currently no PhD program in the department. He said
that they have proposed one, but that the Ministry of Education (Monbusho)
is balking at the cost despite the current shortage of Ph.D.'s in
computer science.
After this brief introduction, we adjourned to a lecture room on
another floor for my talk. He had indicated when we were setting up
the visit that a research talk was probably too specialized, so I had
volunteered to give an "Introduction to Distributed Systems" talk based
on a lecture I had given earlier in Tokyo. The turnout was impressive --
about 50-60 people -- of whom the vast majority were students.
Following the talk, he spent some more time showing me around the
department as we talked more about his research, the department, and
the university in general. His basic work is involved with pattern
recognition algorithms, specifically, the construction of document
recognition systems. He had his students show me a demo of the
latest system, which will convert pages from a scanned scientific
document into TeX input. The systems, which is written in C for Sun 3/4
workstations and has an X-windows interface, deals with both roman and
Japanese characters, tables, scientific equations, and figures. I
certainly not an expert in such systems, but it compared quite
favorably to the commercial systems that I had seen. He gave me a
copy of a paper that he had just returned from presenting at an
IEEE Document Recognition Workshop in New Jersey. He mentioned also
that people at Bell Labs are working on such systems and are perhaps
the leaders in the field. By the way, the laboratory itself contained
a fairly impressive array of equipment, including Suns, NEC PCs, and
an Omcron Luna workstation.
Following the demo, he showed me some of the departmental instructional
machines. The "big iron" here was a Convex that runs Unix and can support
up to 200 undergradate users simultaneously. The machine was bought
3 years ago using money provided by Monbusho, who also supplies money for
maintenance. In response to my query, he says that Monbusho will, as a
rule, give money for such equipment about once every 10 years.
Since this time period greatly exceeds the life span of most computer
equipment, they have to be careful about being stuck with obselescent
equipment. The strategy they use at Shinshu is to lease the equipment
so that it is more easily replaced. When I asked about why they
had chosen a Convex, he indicated that the availability of NFS at the
time of purchase was one important consideration.
We then went up on the roof to admire the view (the building is one
of the tallest buildings in Nagano.) He also showed me the tennis
courts next door which are scheduled to be replaced with a new building
within the next few years. The building be about 50% of the size of
their existing building, and is being constructed mostly to help
accomodate their increased enrollment. Since Okamoto enjoys tennis,
I got the feeling that he had decidedly mixed emotions about the whole
thing!
Upon returning to his office, he gave me a brochure describing the
department. Although it is in Japanese, it does contain an understandable
picture of the departmental network. There is also a university network
developed largely by the department. They are connected to JUNET through
a line to an Epson facility in Matsumoto, where the main campus of the
university is located.
I came away impressed with the situation at Shinshu. For a university
at the top of the 3rd tier in Japan (as Okamoto characterized it), the
equipment and general Monbusho support seem better than I had expected.
I was also impressed that Okamoto himself maintains an active research
program and is successful in publishing. I frankly can't think of many
people at equivalent universities in the U.S. that do as well. While his
kind of success may not in fact be the norm, my overall impression is
that there is perhaps less distance between the best universities and
universities such as Shinshu than one would find in the U.S.
----------------------------
Prof. Masayuki Okamoto
Dept. of Information Engineering
Faculty of Engineering, Shinshu University
500 Wakasato, Nagano 380
Japan