第19回OCLC研究図書館館長国際会議とOCLCワークショップにおける愛知淑徳大学図書館の報告

Thank you Mr. Wang, and Good afternoon ladies and gentlemen.

My name is Hiroshi Itsumura. I am an Associate University Librarian of Aichi-Shukutoku University Library and I am also a member of the faculty in the Department of Library and Information Science.

Before I begin, I would like to thank Mr. Wang and OCLC for extending their Invitations to me, and also to my colleague and to our Graduate students, to attend the OCLC's 19th Annual International Conference of Research Library Directors.

Here is the statistics as of year 2000.

As you can see in the screen there are 99 national, 70 public, and 560 private universities in Japan. The number of volumes owned by the libraries totals to over 3 hundred 38 million. The budget totals to over 7 hundred eighty million US dollars.

The biggest bibliographic utility used in Japan is the NACSIS-CAT system developed by the National Institute of Informatics, which is a part of Ministry of Education(Now officially called the Ministry of Education, Culture, Sports, Science and Technology)

Most commonly used classification scheme is NDC, the Japan Decimal Classification and DDC and LCC, but there are 18 standard classification schemes to choose from. Over 95% of University libraries use NDC now. The level of description is left up to each participating libraries.

There are of course libraries in Japan that use OCLC's WorldCAT as well. I heard that there are about 50 (http://www2.oclc.org/cgi-oclc/pi.pl) member libraries in Japan, but I think Mr. Wang can give you more accurate figure than I.

There is so far no utility such as CORC that can handle metadata, in Japan.

Next, I would like to talk to you about Aichi-Shukutoku University and the Library.

As you can see, Aichi- Shukutoku was first founded as Aichi Shukutoku School for Girls in 1905.

In 1961, Aichi-Shukutoku Junior College for women was established.

In 1975 a 4-year College for women with Departments of Japanese and English Language and Literature was founded.-The library table 3

The University has two campuses and there is one main library and a branch library.

The total monographic holdings is 260,000 volumes and we currently subscribe to 1,059 journal titles (447 western language, 583 Japanese, 29 Chinese). The rest is western language.

The budget is $1 million.

The current student enrollment is 7000, of which there are 200 in the Graduate programs.

As you can see from this chronology, our involvement with CORC took on a form of a cooperative effort between the Library and the Department of Library and Information Science. Since CORC was a very new concept and there were almost no literature written about it in Japanese at the time of our involvement, we felt that this was the most feasible way for us to take our first step forward. It was probably both realistic and practical for a small library such as ours to convince somewhat conservative? Library Board Members to understand the Library's role in such an endeavour.

Next, I would like to point out some of the difficulties we have had, and also issues we would like to do further research on, and our hopes for the future development of CORC.

Problems unique to JAPAN

The Major problem for most of us Japanese is our proficiency in English. Because all CORC manuals and tools are written only in English, we feel that many Japanese librarians will find it difficult at first to attain the necessary skills to work with CORC. It will be a great help if the manuals were to be translated into Japanese. Although I do understand the difficulty as CORC is still evolving, and manuals are constantly being revised in the process.

As I have mentioned to you earlier, most Japanese librarians are not familiar with USMARC format with the exception of OCLC member libraries, of course. As many literature on the subject have pointed out, USMARC is very detailed and requires many years of training for one to work with confidence. There isn't a published manual in Japanese that explains in detail how it should be used. The National Diet Library, which is the Japanese equivalent of the U.S. Library of Congress, uses JAPANMARC, which was developed from UNIMARC as its base. Unfortunately, because JAPANMARAC was developed specifically for the needs of the Diet Library only, it is not used widely among academic libraries.

AACR2 also is not as thoroughly understood and used by many libraries in Japan, as one would expect them to. This is mainly due to the fact that only the 1978 edition of AACR2 has been translated into Japanese, and later editions are only available in English.

With rapid decline in birthrate, the student population has also decreased in the past 10 years in Japan. As a result, many universities in Japan are facing financial strain and have had to cut back on their budgets. The library budgets, especially in terms of personnel are most hard hit. Private universities such as ours, or at least not too many of them, can afford the luxury of allocating multiple number of staff on projects such as CORC. Mizuki Kashima our librarian has been the only staff to work with CORC since the beginning and she has had to divide her time among other duties in the library.

I do envy some of the American libraries that appear on the CORC Users Group Home page, where a group of librarians are working together to form a team to work on CORC.

So far, there has not been any national consensus on the future of metadata in Japan. Our bibliographic utility NACSIS-CAT has yet to announce its policies on its handling of internet resources.

NACSIS-CAT has recently added IDENT field to encode url in the bibliographic record for e-journals that are accessed remotely via the internet but it has not yet addressed the issues of other structural metadata that are needed in monograph-type resources.

The National Diet Library has recently announced that it is in the process of planning a project to create metadata for internet resources that it deems valuable to be archived and that they have decided on using Dublin Core as their metadata scheme. They say that they plan to start the Project in the Spring of 2002.

This is not only unique to our situation, but because CORC does not yet support the use of Japanese characters in the description, we could only catalog resources that were written in English. We do believe that there are many valuable resources that are available via the internet that are only written in Japanese. We are anxiously awaiting for OCLC to support CJK characters in UNICODE. NACSIS-CAT has recently developed its new Cataloging system to support UCS(Universal Character Set), which is a subset of UNICODE so that both Chinese and Korean vernacular scripts can be expressed in our bibliographic record. Many Library system vendors, as a result, are developing new systems based on NII's decision to implement UCS. I would be most interested in hearing from other Asian libraries how they have tackled the problem of vernacular scripts.

There are several very important issues we must look into, in order for us to investigate the realistic application of metadata in our libraries in Japan. A few national universities have been developing digital libraries. Among them, University of Library and Information Science, which utilizes DC to create metadata for constructing its subject gateways. We too, need to further examine the use of Dublin Core as an alternative to bibliographic description of internet resources. In order for us to do that, we will need to study more closely possible mappings between DC and NACSIS-CAT format, and to define our metadata registry that best suit our true needs. Because metadata is still a very new concept in Japan, and compared to the business world libraries are also slow to catch up, we feel that having had the firsthand experience with CORC, we must be prepared to make recommendations whenever we can.

It goes without saying, that without a well thought out policy for selection of internet resources, half the effort in creating metadata will be wasted. We have yet to create a workable policy for selection and evaluation of internet resources and it will be our next project to have the reference librarian take part in such a program.

We will need to follow closely OCLC's policies as well as the Library of Congress's guidelines and policies on selection of internet resources.

In order to expand on the possibilities of adding Japanese language resources to CORC and to export such records to our local library system, we are again looking towards OCLC and CORC to come up with solutions to make all of these possible.