MIT
Reports to the President 1994-95
At MIT prototyping is defined as that development area that links research and
operational services. For the Libraries and Information Systems, the
Distributed Library Initiative (DLI), has provided a prototyping "tank" where
promising projects can be explored, exercised, and evaluated. By working with
researchers, new ideas can be investigated, and by working with libraries'
staff, old services can be renewed and new solutions can be explored for
existing problems. This prototyping space fits the MIT style very well because
its goal is to provide a functional capability that enhances services. This
year, the Libraries pursued a variety of prototyping projects.
Agreement and working activity with Geac Computers, Inc. on co-development of
next generation, client-server based library systems.
As described in the new system description above, a key component of the MIT
agreement with Geac Computers is a section of the contract which specifies the
nature of our Co-Development agreement. MIT would rather work with a vendor on
the design and distribution of a product than receiving a "finished" product
that does not meet our needs. To remedy this handing-off scenario, MIT
actively works with vendors such as Geac to define requirements jointly and to
work with the vendor as an expert customer in order to deliver the best product
possible in the most efficient amount of time.
The Geac/MIT Co-development agreement specifies MIT involvement in all areas of
a library automation system, including new Web-based services, digital image
delivery, archives management, and security and networked infrastructure
requirements. MIT staff have worked actively with Geac developers and with
other interested Geac sites such as NYU, SUNY/Albany, and Los Alamos National
Laboratory Library to provide a full client/server library system by the summer
of 1996.
TULIP stands for The University Licensing Program, a consortial experiment
sponsored by Elsevier Science Publishing. At MIT, the goal of TULIP has been
to place online page images of Elsevier journals in the field of Materials
Science. This experiment has involved IS development staff and Libraries'
reference staff to build and test a prototype architecture to deliver image
based pages of scholarly journals. In TULIP, students, faculty, and
researchers can search files via a bibliographic search client - Willow,
developed at the U. of Washington and augmented with significant national
standards programming of Z39.50 by MIT. TULIP is also accessible via a Web
server which permits access by journal title, issue, and table of contents.
The image browser for the Tulip images, developed by Bill Cattey of DCNS/IS is
a powerful, richly functional tool that provides very fast access to the TULIP
journal page images and supports the functionality to browse through articles,
magnify pages, print, and further manipulate the images according the needs of
the user. The Libraries continue to acquire the necessary storage to provide
access to TULIP; currently the datasets cover almost three years of the 38
journals in the MIT TULIP set. The system supports an anonymous logging
mechanism to provide quantitative data regarding use of the journals; this
logging does not capture any personal information about a user. It does gather
data about the journal title, journal article, use - print or view, etc. The
TULIP implementation team has met frequently with representatives from Elsevier
Science Publishing to show them the system, and, led by Suzanne Weiner of the
Libraries, the team has initiated an outreach program to publicize the system
and to encourage members of the Materials Science community to try this system
and to provide us feedback.
For several years the Libraries have operated small, localized LANs to provide
efficient access to CD-Barton and to support some local operations. LAN
services across the entire Library enterprise, however, have not been feasible
because of policies regarding network protocols for MITnet and proper naming
and security tools. This year Ray Charbonneau led an effort to experiment with
a new version of Novell Netware IP to extend local network services across
MITnet. He worked with IS Distributed Computing and Network Services and
received training in this new product. At year's end the experiment was
successfully supporting CD-Barton at Branch Libraries, including circulation
status. Based upon Library purchasing and licensing decisions, other CD-ROM
services could be provided, and discussions are underway regarding efficient
distribution and licensed use of administrative and office software that would
be provided by the LAN server.
MIT is a beta site for WebZ, a new product developed by OCLC that combines the
ubiquity of Web browser clients with the Z39.50 protocol-based searching of
networked library catalogs. Tom Owens is heading the effort at MIT to install,
test, and develop Web-based search interfaces for Z39.50 queries. This beta
effort will end in July 1995, and could become an important component of the
Libraries' new catalog environment. Via WebZ, MIT and other internet users
could link to a Web page to search the MIT online catalog. Because of the
universal availability of Web clients, this will alleviate the effort to
maintain specialized online catalog clients, and it will provide a search and
retrieval standard that many vendors and library catalogs support.
The Libraries have been building an electronic reserve module residing in the
Athena environment. This would provide electronic access to reserve materials
thus saving student and faculty time and effort to build a paper reserve
collection. The Libraries' prototype for electronic reserves builds upon the
architecture created for TULIP and the conversion of paper-based information
using knowledge gained as part of the CS-TR project described below. By
building upon these existing tools and architectures, the Libraries have been
able to create an electronic reserve library system. The first area of focus
is the exam "bibles." The Libraries have begun to experiment with scanning the
pages of the bibles and to make them available via the TULIP system structure.
Libraries' staff attended a national seminar on electronic reserves and the
technical, service, legal, and curriculum issues this effort represents. A
sample file of electronic images has been built.
As with many experiments, the next logical steps of moving to production are
impeded by the issues of intellectual property. The Libraries have begun the
reserve efforts based upon materials created at MIT. Even these, however, have
copyright issues and permissions which must be addressed.
The Libraries' Systems Office and MIT Archives have created a successful
implementation to place the minutes of MIT Faculty Meetings online. Lois
Beattie of the Archives successfully converted the minutes to electronic form,
and Carter Snowden of Systems created a Web environment to browse and select
the minutes. This project was announced to the Faculty Committee by chair
Robert Jaffe at the last Faculty meeting of the year.
Tom Owens of the Library Systems Office is participating in an effort to
provide a coherent structure to development efforts in the MIT Delivery
Process, a key area in the new IT framework on campus. The goals of this new
development process are to provide improved productivity, more consistent
levels of quality products and services, and to foster a better understanding
between customers and developers.
In order to anticipate and incorporate new technologies, the Libraries must
maintain close working relationships with appropriate research groups at MIT
and elsewhere. At MIT, the DLI has provided a common ground for researchers,
librarians, and information technologists to meet to explore new areas and to
plan for future library systems. The synergy in this environment is highly
productive: the Libraries provide content and information management
experience; IS provides scalable technology for delivery and supportable
infrastructures; research groups such as Professor Jerry Saltzer's Library 2000
provide the critical thinking of research and testbeds to explore the myriad
components of the digital library.
The Libraries have benefitted enormously from the contributions of Professor
Saltzer. His world class thinking and analysis have moved the Libraries' and
IS work forward, and he has served as a valued colleague and mentor for many.
As he scales back his academic and research agendas this year, the Libraries
must work to build similar relationships with other researchers at MIT.
This project has been described in previous annual reports. As part of a
national consortium MIT is participating in the construction of a digital,
distributed library of computer science technical reports. The primary
contribu- tion of the MIT Libraries has been located in Document Services where
a large-scale operational scanning effort has been built. Document Services
has now reached a production level of over 1000 pages per week of high-quality
scanning. These images are then transferred via FTP to Library 2000
headquarters in the Laboratory for Computer Science where they are made
available. This production environment is the result of intense, focused
effort to investigate, analyze and improve every step in the process from
selection and gathering of the paper documents, to preparation and scanning,
quality-control, transfer of the images, and storage of the raw files.
As the project funding comes to an end in FY1996, efforts have begun to move
the experiment into the Libraries' operations and to continue and expand the
availability of the content.
As members of the Boston Library Consortium (BLC), the Libraries participate in
the shared technology program of the member institutions. This year the
Information Technology Committee of the BLC proposed a grant to provide Z39.50
support for all participating members. At year's end this grant was approved,
and MIT will receive its share to help support Z39.50 components in the new
library operations system.
As the year closes with the recognition of these achievements, the Libraries
must now consider whether it can sustain and support all of these efforts and
any new efforts that may present themselves in the coming year. In the coming
year these are some of the critical issues which will face the new Library
Director in the area of library information technology.
In order to continue and solidify the impressive developments of the DLI, the
Libraries and IS must allocate additional staff for development, operations,
and sustainable support for these systems. The Libraries must find additional
financial resources to fund electronic services that are in addition to
traditional services and to provide the support for continued improvement and
development of new and desirable services. Staffing in Library Systems is a
critical issue. Although every aspect of the Libraries' organization shares
responsibility for the support and management of the technology base, Systems
has core responsibilities for the infrastructure and operational services. It
should be noted that no new staff have been added to Systems since the
CD-Barton implementation in 1989/90. Since then, more network-based services
have been added, CWIS services have been added, the Libraries have been
networked completely, the Libraries have completely upgraded its microcomputer
base to networked Windows machines, prototype services such as TULIP and a
Library-wide LAN are being moved into an operational environment, and new areas
of work such as the CS-TR project must move to the Libraries this fiscal year.
How will the Libraries and IS provide support for this increasingly extensive
and complex technology environment for electronic library information?
Continuing discussions in the areas of support models and funding for that
support are critical to the DLI.
The Libraries and IS have been working in order to understand the issues
related to capacity planning for an increasing amount of electronic
information. Through the DLI structure this issue must be explored and
possible solutions examined.
In some respects the DLI focus on collaboration and shared involvement and
responsibility between the Libraries and IS provide a few early models for the
new IT framework being implemented at MIT. Building upon the existing DLI
framework, the Libraries have an opportunity to remain a leader in coordination
of work with other MIT departments. The new IT process must not hinder work,
even in the transition period. The Libraries can nurture this transition by
maintaining progress in each of these project areas and by continuing to be
open and available as a credible partner in the IT arena. The new Director
will need to ensure that the principles and relationships in the DLI realm are
maintained both in the Libraries' leadership transition and in the new IT
framework at MIT.
I am an optimist, and I believe that MIT is the obvious and best crucible for
an intensive and extensive exploration of all aspects of digital libraries.
The research interests at MIT are consistent with the information age focused
on the network, information delivery, and services. Vendors are interested in
working with MIT because of the knowledge and perspective based in the
Institute, and MIT staff have the expertise and experience to realize the
promise of digital libraries. MIT has a distinguished research record in many
disciplines; it has a history of interest and exploration in library
information; therefore, it should devote the resources and commitments to
becoming the leader in digital libraries. I believe that we are poised to take
that step and to enter a golden age built on the established traditions of
collaboration, openness, and creative initiative. It has been my pleasure and
great good fortune to have been in the Libraries during this highly creative
and stimulating period. It has been and will continue to be my good fortune to
work with MIT staff who define excellence and heroism every day.
Greg Anderson
MIT
Reports to the President 1994-95