- page 1 - Anna Gold - 7/4/12
SDSC - UCSD Libraries
SDSC Assessment of Information Needs and
Opportunities for Partnerships
2. Goals of the Assessment
3. Interview Methods
4. Interview Findings
4.1 Needs Assessment: Initial Survey
4.2 Needs Assessment: Briefing
4.3 Needs Assessment: Structured Interviews
4.4 Needs Assessment: Unstructured Interviews
4.5 Research Collaboration: Unstructured Interviews
5. Analysis of Findings
- page 2 - Anna Gold - 7/4/12
SDSC - UCSD Libraries
SDSC Assessment of Information Needs and
Opportunities for Partnerships
GENERAL: The San Diego Supercomputer Center (SDSC) is a unique and nationally
reknowned center for scientific computation and supercomputing. Founded in 1985, in
the last dozen years it has changed significantly both in size and mission, going from
fewer than fifty employees in the late 1980’s to an organization of over two hundred and
fifty employees in 2000. It has recently become an organized research unit (ORU) of the
University of California, San Diego. While some of its staff are affiliated with and teach
in other campus departments, the majority of its staff are non-teaching researchers,
administrators, and computer scientists who support the extraordinary computational
resources of the Center.
MISSION: Formerly one of a handful of national supercomputer centers, in 1997 the
SDSC became one of two national centers or “leading edge sites” in the NPACI (National
Partnership for Advanced Computational Infrastructure), currently consisting of 40 U.S.
and 6 international affiliates. The other leading edge site is the National Center for
Supercomputing Applications (NCSA) Alliance, in Illinois. In addition to these two
national sites, SDSC’s peer organizations include major supercomputing operations at
federal laboratories such as Los Alamos National Laboratory; agencies such as NCAR
(National Center for Atmospheric Research); and regional supercomputing facilities such
as the University of Pennsylvania’s.
As an NPACI leader, SDSC works to establish, through hardware acquisitions and
software development and integration projects, a national scalable, high-performance
computational infrastructure to support academic research. In support of this mission, the
- page 3 - Anna Gold - 7/4/12
SDSC houses an array of powerful supercomputing resources appropriate for use in
computational science, and is responsible for allocating time on these machines to
researchers around the country1. The SDSC also directly supports a number of research
programs aimed at advancing the state of infrastructure for computational science.
Computational science areas with the most active research programs at SDSC are
computational biology and computational chemistry; important infrastructure research
projects include work on curating data, and computational science workbenches and
In view of SDSC and NPACI missions, support to remote users occupies a central place
at SDSC. This support includes outreach and training on the use of SDSC resources in
the form of workshops, tutorials, and a variety of other conferences and events. Impact
on and benefit to the research community are the primary measures of SDSC staff
effectiveness. The SDSC differs in this respect from more traditional department-based
research. SDSC staff have relatively few curriculum-driven teaching responsibilities and
relatively little graduate student assistance. Their research efforts are designed to create
robust infrastructures of immediate actual utility within the community, rather than tools
whose primary goal is “proof of concept.”
STAFF: Personnel at the SDSC are comfortable with sophisticated technology and
highly diverse in terms of their technical and scientific expertise. The number of SDSC
personnel continues to grow rapidly to meet the demand for program support; in the last
two years the number of SDSC personnel has exceeded the capacity of the primary SDSC
building. A handful of SDSC staff are now located in offices elsewhere on the UCSD
campus, and a small number have offices located off campus. As a consequence of this
rapid growth, space at the SDSC is at a premium; many individuals at SDSC now share
offices with one or two other colleagues.
1SDSC hardware ranks 16th, 116th, and 262nd on a popular list of the world's 500 most
powerful computers in July 2000 (http://www.top500.org).
- page 4 - Anna Gold - 7/4/12
SDSC FUTURE: The SDSC will continue to add personnel and will require new
building space just to adequately house their existing staff. Over the next several years,
SDSC administrators anticipate fostering a close relationship with the new CAL- IT2
Center (California Information Technology and Telecommunications <?> Center), which
could include developing shared information or library facilities, services, or resources.
IMPACT OF THE DIGITAL INFORMATION ENVIRONMENT: As in all areas of
science and engineering, the information environment for scientific computing and
supercomputing has evolved rapidly in the last ten years. The Internet has made informal
digital scholarly and professional communication the norm; email, listservs, and personal
web pages now provide a suite of informally administered tools2 on which scholars rely
heavily, supplementing rather than supplanting face to face communications. In addition
to this transformation of personal communications, an enormous proliferation and
transformation of formal and semi-formal digital publishing has taken place3. Formal
communications are increasingly handled in digital formats, with electronic submission
and review of journals and accelerated turn around times for publication4. Familiar tools
of scholarly communication in their digital formats are well-understood and embraced
with enthusiasm. Yet researchers face growing difficulties in navigating this expanding
universe of content. These burdens have been partly alleviated by increasingly accurate
2E-mail, the original Internet "killer app," is ubiquitous and supported by numerous
commercial as well as open source applications. Listservs, newsgroups, and electronic
mailing lists support loosely affiliated communities sharing common problems and interests.
They range from open to closed lists and are supported by a wide range of simple commercial
or open source applications. For academics, personal web pages have become almost a
requirement of doing business. They are the electronic equivalent of a business card or
directory listing, but for many they also represent an unwanted maintenance problem.
3 At UCSD more than half the journals subscribed to by UCSD Libraries are now available in
digital editions. Electronic books are now beginning to penetrate academic as well as
popular markets. Electronic preprints have largely supplanted their print predecessors;
technical reports, conference papers, government reports, proposals and funding documents
are widely expected to appear in electronic versions, and many are available only
4 Numerous scientific professional societies, in mathematics, physics, chemistry, and other
fields now accept submissions electronically and handle peer review electronically. The
National Science Foundation and other federal agencies are rapidly moving to electronic
proposal and review processes.
- page 5 - Anna Gold - 7/4/12
popular free search tools (the current hands-down favorite appears to be Google at
http://www.google.com); but formidable problems remain in navigating and finding
information effectively. Scientists and professionals do not have (or do not think they
have) the time to learn how to make effective use of new information interfaces and tools,
and many rely on “tried and true” but sometimes limited or outdated techniques to
SCIENCE ENVIRONMENT: At SDSC, the digital environment and the application of
computational techniques to scientific problems have encouraged an interdisciplinary
style of research and work that draws together fields such as mathematics, computer
science, chemistry, biology, physics, and engineering. The diverse scientific
backgrounds of the SDSC staff makes SDSC a kind of “university within a university”
where a researcher with a need to know something about a new discipline can in all
likelihood turn to a colleague elsewhere in the building for expert guidance.
PREVIOUS LIBRARY SERVICES AT SDSC: Until about two years ago, SDSC
supported a small special library. At one point the library occupied a large, centrally-
located office space in the main SDSC building. Staff assigned to the library included a
mix of part-time paraprofessionals and professionals. When it existed, one of the most
valued contributions of the SDSC library was its support for maintaining a database of
scientific research accomplished by people who had been given access to SDSC
resources (that is, time and space on SDSC machines). This database was occasionally
published as a bibliography, most recently in 19916, and it was last updated as a database
in fall of 1996. A small collection of publications, including manuals, newsletters,
videos, and books, was managed via a second database acting as a library catalog. After
an SDSC librarian resigned in 1998 to accept a full-time position with UCSD Libraries,
5It might be claimed for example that the browser wars have subsided in part because users
are not willing to spend time working with unfamiliar tools such as Opera, a multi-screen
browser that adds great flexibility to the small-screen web user’s experience.
6Computational Science at the San Diego Supercomputer Center: A Bibliography, July 1991.
Edited by Mary Layman, SDSC Librarian.
- page 6 - Anna Gold - 7/4/12
the library space was converted to badly needed office space, and its collection was put in
storage. At about this time the small hard copy collection and its catalog, and active
maintenance of the research bibliography, were all suspended.
2. Goals of the assessment
In July 1999, the SDSC asked UCSD Libraries to provide an analysis describing what
information services SDSC might enjoy if they were to reestablish a half-time librarian
position. UCSD Libraries responded with a proposal to provide a two-part assessment.
The first part of the assessment would address the current state of SDSC information
needs, in view of recent changes in scholarly and technical communications, and the
growth of digital e-journal availability. The second part of the assessment would address
the potential for digital library research partnerships between SDSC and the UCSD
The primary goal of the assessment was to provide a strong foundation on which future
SDSC library and information services might be built: one informed by the variety of
staffing and service models that might be appropriate, and flexible enough to adapt and
respond to continued change brought on by rapid evolutions in digital scholarly
communication and digital libraries. The secondary goal was to establish a basis for
ongoing communications, mutual support, and co-development of digital libraries
research between the SDSC and the UCSD Libraries, in order to take advantage of
synergies between the two institutions. Most important, it was felt that this secondary
goal would provide increasing support for the primary goal by adding the many
dimensions of digital library development to the effort. Through this process, a mutual
understanding and knowledge of complementary capabilities, interests, and goals of the
two parties will be gained.
In March 2000, SDSC and UCSD Libraries agreed that UCSD Libraries would provide
the assessment described, and in April the project was outlined and scheduled. The plan
for the assessment consisted of the following elements:
- page 7 - Anna Gold - 7/4/12
3.1 An e-mail survey of all staff which invited their answers to a number of questions as
well as their participation in informational interviews. (Completed April 2000)
3.2 A staff briefing, intended to provide a general update to interested staff on relevant
developments at the UCSD Libraries. (Completed April 2000)
3.3 A structured interview conducted with volunteers and a number of key informants
who were identified through the interview process (the "snowball" technique).
(Completed June 2000)
3.4 An unstructured interview combined with reactions to a model digital library
scenario, conducted with a second set of volunteers and key informants. (Completed
3.5 In parallel, several unstructured interviews with SDSC digital library researchers to
help elicit and explain options for digital library collaborations between UCSD
Libraries and SDSC. (Completed June 2000)
3.6 A draft report on information needs at SDSC incorporating a draft report on potential
partnerships between UCSD Libraries and SDSC, to be shared with key informants
at SDSC and UCSD Libraries. <Completed August 2000, revised September 2000, to
be shared with SDSC staff on acceptance, fall 2000>
3.7 A final summary report on information needs at SDSC, incorporating SDSC and
UCSD Libraries feedback and additional information needed in order to implement
the report's recommendations. <Draft completed September 2000 >
4. Needs Assessment
4.1 INITIAL SURVEY: A survey based on questions developed at MIT for a study of
faculty information needs was developed and sent by email to all staff at SDSC (see
appendix 1). The intent of the survey was not to gather definitive data but to help
inform the construction of the interviews to follow, as well as to solicit an initial
group of volunteers for interviews. Eight surveys were returned, and while the results
were therefore far from definitive, the results conformed well to the (later) interview
- page 8 - Anna Gold - 7/4/12
results. Among the results were indications that the most frequently used sources of
information (used at least weekly) at SDSC were:
Results of Internet searches (8)
Known / bookmarked web sites (8)
Personal contacts (7)
General interest newspapers and magazines (7)
Electronic journals (6)
Internal SDSC databases or publications (6)
Less commonly used sources of information were:
Printed journals subscribed to personally or by SDSC (4)
Books bought personally or by SDSC (4)
Desk copies of manuals or handbooks (4)
Least frequently used information sources were:
Printed journals at UCSD Libraries (2)
Books from UCSD Libraries (2)
Preprints and technical reports (2)
Conferences and proceedings (2)
Workshops, courses (2)
Numeric data (2)
The very least frequently used information source was an information specialist or
librarian, with only one respondent selecting this as a "frequent" source of
The survey respondents had a number of different ideas about what information tools
they would like to have that did not exist; five indicated interest in tools to help
manage or access information about SDSC: a bibliography of SDSC published
research; a bibliography of research that benefited from SDSC resources; internal
databases at SDSC; a database of SDSC public relations information; or better access
to SDSC internal documentation. When asked to describe their ideal on-site
- page 9 - Anna Gold - 7/4/12
information facility or library, three indicated "none" - that is, they preferred
accessing information through the web rather than in an on-site facility; a fourth
commented that they wanted access to a librarian, but not to an on-site library.
4.2 BRIEFING: On April 6, former SDSC librarian Laura Smart conducted a one-hour
briefing at the SDSC general auditorium to which all SDSC staff were invited. It was
attended by 22 staff from several different areas of the organization7. In her
presentation she emphasized access to electronic journals, an area which had
progressed rapidly in the previous two years; and also new research tools such as
cross-database search programs (Database Advisor or DBA, developed at UCSD
Libraries, and SearchLight, the California Digital Library's implementation and
adaptation of DBA). Another element of her presentation was showing how to use
the Science and Engineering Library's web site to access these and other tools and
services, including a database of selected web sites and web-based research sources.
As an unplanned but welcome follow up to this briefing, UCSD Libraries staff
involved in the assessment (Anna Gold and Laura Smart) were invited by Jay
Bouisseau, Associate Director for Scientific Computing, to provide a similar briefing
to his group at a regularly scheduled staff meeting on April 26. Anna Gold provided
a summary of the points covered by Laura Smart, and also prepared a number of
handouts and a web page with links to electronic journals and other library services
of particular interest to SDSC (appendix 2). Follow-up questions raised questions
about what options were available for borrowing and accessing materials; a number
of participants voiced the opinion that a good one-stop web site customized to
provide access to most frequently needed services and resources would be the most
useful service UCSD Libraries could provide.
4.3 NEEDS ASSESSMENT: STRUCTURED INTERVIEWS: Next, a structured set of
questions was developed (appendix 3), and appointments were arranged with five
- page 10 - Anna Gold - 7/4/12
volunteers and key informants representing a variety of functional areas at SDSC
(scientific research, outreach, scientific computing), whose tenure at SDSC ranged
from a few months to fifteen years. Two had external affiliations which involved
them in teaching responsibilities.
The first set of interview questions concerned the participants' areas of expertise:
recent changes in their field; the impact of interdisciplinarity on their field; what was
unique about their work at SDSC; their own methods for keeping current in their
field; and what the important means of peer communications were in their field.
RECENT CHANGES: A number of interviewees pointed to the impact both of
computational power and of the web as a distribution medium. On the one hand the
"web has changed everything" about the way scientific and technical
communications are transacted. A different trend has been that huge quantities of
data are now available, along with capabilities for large storage and fast processing;
yet making this data retrievable and usable remains an enormous challenge. The
growing significance of computational methods across numerous disciplines was also
noted as an important trend.
INTERDISCIPLINARITY: Among the many disciplines represented at SDSC are
numerical and applied mathematics; computer science; chemistry; biology;
environmental science; education; networking technologies; some physics; and
increasingly the social sciences and humanities. SDSC differs from the other
"leading edge" NPACI site in emphasizing the biological sciences and chemistry
where NCSA emphasizes physics.
WHAT IS UNIQUE AT SDSC: As a national resource, and because it was
administered earlier in its history by private industry (General Atomics), in the recent
past SDSC led a relatively separate existence from other UCSD campus departments
7Staff who attended were from Scientific Computing, External Relations, Support Services,
computational science, DICE, the Visualization Lab, and other operational and program
- page 11 - Anna Gold - 7/4/12
and programs. This is no longer the case: as of 1997 SDSC became an organized
research unit (ORU) of the University; at the same time, SDSC is making an effort to
become more integrated with campus research and academic life. Still, SDSC differs
from traditional research departments in important respects. SDSC has a strong
national service and outreach mission that informs the work of its scientists as well
as its computer professionals and administrative staff; the absence at SDSC of a
degree-granting instruction program is emblematic of this difference. Another
exemplary distinction is that great deal of proposal writing is done at SDSC, often
with the assistance of experienced professional writers expert in the proposal
process; in many departments and programs, proposals are handled directly by
scientists and faculty. As a result, occasionally SDSC lends its proposal expertise to
other campus units.
KEEPING CURRENT: Regardless of their field, the first set of interviewees agreed
that keeping current in their field was difficult at best, though it could be tackled
through a combination of techniques in which personal contacts (ad hoc and at
conferences, seminars, and workshops) figured prominently. Conferences and
seminars will "never be replaced," said one, noting that face-to-face contact is
irreplaceable, and that new ideas are often evolved in these settings. Listservs,
preprints, and technical reports were considered relatively unimportant by most;
peer-reviewed journal articles fell somewhere in between face-to-face contact and
informal scholarly communications, being used principally in the context of the need
to prepare proposals or publications of their own. A principal investigator might need
to follow a few journals closely (particularly if they sat on their editorial boards);
otherwise a person might choose a few titles to follow regularly: they might
subscribe themselves, or regularly browse the publisher's web site, or sign up for a
publisher’s alert service, or sign up for a journal routing service (formerly provided
by the SDSC library). Even with these methods, "it's almost impossible" to truly keep
current in the literature of any field. One described the challenge as a continual "trial
by fire." Most conduct their own research without assistance from graduate students
- page 12 - Anna Gold - 7/4/12
or other people to whom they can delegate this work. The proliferation of electronic
journals has been a great help; most staff know they can easily locate an e-journal's
web site by using a search engine, and feel they can read more, and more easily, this
way. Other techniques cited for keeping current were subscribing to the HPC Wire
email service (a kind of electronic clipping service on high performance computing
news); regularly visiting technical or academic bookstores; or identifying an expert,
finding their leading papers, and checking out the references they cite. Still it was
clear from the interviewees that personal contacts continue to be important to
LEARNING NEW FIELDS: Outside one's familiar territory, learning about a new
area or new field is something that SDSC staff must do frequently. To do so they
often turn to their colleagues within the SDSC, recognizing that a wide range of
expertise is close at hand (and not just among the scientists, but also among SDSC's
writers). In this they feel they enjoy an advantage over some of their colleagues in
more traditionally defined disciplinary research units and departments. They also use
the web to seek out information about new fields, and ask questions of contacts via
email. Yet as one informant said, the interdisciplinary nature of their work makes it
"ineffective" to use traditional databases8, and the effort required to learn about new
fields can be "overwhelming."
A second set of questions concerned how the respondents go about using and seeking
information: whether they use published or unpublished materials; article databases;
browsing; they were also asked to describe how they would go about accomplishing
a number of common information-seeking tasks.
RESOURCES USED: In addition to core professional journals, informal
publications from peer organizations or funding groups (e-prints, RFP's, reports,
8Many bibliographic databases are still broadly defined by disciplinary topic: INSPEC is
used for physics, Medline for clinical biology, BIOSIS for scientific biology; SciFinder for
- page 13 - Anna Gold - 7/4/12
technical papers) were the most frequently mentioned research materials used in
everyday work. The use of article databases was far from universal; while some
were familiar with disciplinary databases with web interfaces, several "never" used
article databases, while others were using tools that were "tried and true" (e.g. telnet
access to "Melvyl" databases that had been available for many years9). Several
respondents had developed simple mechanisms for capturing search results (usually
into text or flat files) for future reference or manipulation.
ACCOMPLISHING INFORMATION TASKS: The respondents revealed an almost
unanimous dependence on a combination of personal searching on the web, and
personal contacts within the SDSC, to accomplish common information-seeking
tasks. True to this preference for working personal networks to get desired
information, several noted that they would appreciate having a personal contact
(“knowing who to call”) at the UCSD Libraries when faced with a “last resort” type
of question about accessing Library-mediated resources or services. Other
impediments to the success of usual information-seeking techniques included
frustrations with “baroque” web interfaces or unfamiliar web sites, and the difficulty
in obtaining relevant results from some web searches. Most did not think these were
problems that could be solved by UCSD Libraries.
A third set of questions was designed to elicit information about current use of
UCSD Libraries physical and electronic “portals” – book and journal stacks, and web
USE OF UCSD LIBRARIES AND WEB SITES: Most respondents were very
occasional users of UCSD Libraries : they rarely charged out books, rarely visited
any of the Libraries, and rarely used any of the Libraries’ web sites. Several had
obviously not visited the S&E Library since it's move to "Central" (the Geisel
9Many were still unfamiliar with the California Digital Library (CDL) as such. The Melvyl
databases that can be accessed through CDL are those that are locally hosted by the CDL.
Many additional bibliographic research databases are available through publisher- or third-
party web interfaces.
- page 14 - Anna Gold - 7/4/12
Library building) over five years before. A common response for these interviewees
when faced with tasks such as locating the table of contents for a recent journal or
finding articles published by a colleague, was to go to a generic web search engine
and do a keyword search in hopes of finding the publisher's or colleague's web site.
ACCESS TO SDSC INFORMATION: When faced with the task of locating
information about SDSC projects, their responses varied according to their tenure at
SDSC: some knew the names of individuals likely to have institutional memories of
these projects; more recently arrived staff simply said this would be “very, very
difficult.” When asked what library services they had found most valuable in the
past, responses included “librarians” (that is, a person to contact who could help
them) and the SDSC bibliography of research, which had been produced by the
SDSC librarians. SDSC has recently made some progress in its own information
management through implementation of a search capability on its web site
(Ultraseek); while an internal document management system (Docushare) allows
staff to share working documents.
A fourth set of questions elicited information about personal information
management systems used by the respondents; and the feasibility or value of sharing
such systems with their colleagues.
PERSONAL INFORMATION SYSTEMS: Most of these systems in use are
elegantly simple, and closely tailored to the needs and working styles of the
individuals: some use small filing systems or diaries; others rely on their
workstations’ file directory systems; others on a tightly edited file of web
bookmarks. While several feel that there would be some benefit to sharing their
bookmarks files with their colleagues, others question whether it would be possible
to identify resources of broad interest and application given the diversity of
information needs across the Center.
- page 15 - Anna Gold - 7/4/12
In a final set of questions the respondents were asked to assign a priority (high,
medium, or low) to a number of generic library and information services: alert
services; convenient access to library materials; research help; one-stop web access;
and better access to SDSC institutional information. The services ranked “high” by
more than one interviewee were the following:
o Desktop electronic delivery of Library-owned or ILL articles (4 High, 1
o Help managing or accessing SDSC-produced publications and information (3
High, 2 Medium or qualified priority by expressing concern about
o Alert, notification service for new web sites or library services in your area (3
High, 2 Medium)
o Alert/ notification service for new research in your discipline (3 High, 1
Medium, 1 Low)
o Pick up and delivery at SDSC of Library-owned physical materials (2 High, 2
Medium, 1 Low)
Other services that were ranked "high" by at least one interviewee were:
o Personal advice or hotline to get help using Library databases or services (1
High; 4 Medium)
o Special or custom web site for SDSC for important and useful information (1
High; 4 Medium or qualified priority by saying SDSC should already be
o Help sharing library / research recommendations with colleagues (1 High, 4
Medium or qualified priority by expressing concern this would be redundant
with existing tools)
o Alert / notification service for issues of general science publications like
Science (1 High, 3 Medium, 1 Low)
- page 16 - Anna Gold - 7/4/12
The respondents were also asked to suggest any other library or information services that
they felt would be worth developing, that had not been listed by the interviewer. Among
their suggestions were:
o Developing a capability of sharing symposia in a cross-campus archive open to
and inclusive of all UCSD departments.
o Adding an electronic reference link to the Libraries’ home page, with an
immediate “we received your email” response to improve customer satisfaction.
4.4 NEEDS ASSESSMENT, UNSTRUCTURED INTERVIEWS: A set of seven
unstructured interviews was then conducted with another group of SDSC personnel,
again representing a mix of administrative, outreach, scientific, and computing roles,
tenure, and location in the organization. These interviews included two staff members
whose offices were in SDSC’s off-site locations (Sequoya Hall and Chemistry Research
Building, each about 10-15 minutes walk from the SDSC main building).
The respondents were asked first to describe what they felt were ideal library or
information services in each of three categories: information system; personal expertise
and assistance; physical space and collection.
INFORMATION SYSTEMS: A system that was repeatedly mentioned as having high
value to SDSC was a system to track research done with SDSC computational resources.
The absence of a tracking tool was felt to be a significant gap and a “big problem;” when
asked to respond to inquiries about the impact of SDSC efforts on computational
science, there are no systematic tools available on which to base an answer. While this
is an important mission-related and administrative need, most also felt that the
compilation of such a bibliography or database had the potential of being an important
research tool, particularly if a system developed at SDSC could be extended to
incorporate citations to research done at the other NPACI sites, and / or made available
to the larger computational science community.
- page 17 - Anna Gold - 7/4/12
Having such a resource would help researchers avoid duplication and identify potential
synergies in their research, particularly in view of the highly interdisciplinary nature of
their work. One respondent felt that this effort was substantially different from the
development of library services; another felt it was a quintessential SDSC library
function. Manufacturers of hardware used at SDSC also have an interest and a stake in
resurrecting and publishing this database, and may be willing to contribute funds toward
this task. Because the NSF expects SDSC to maintain this data, they would be unlikely
to allocate additional funds just for this task; however they might be interested in
supporting a more ambitious project encompassing citations from multiple national and
international centers of research, in anticipation of its bringing clear research benefits to
a larger community. The special value of such a tool is that most scientific publications
tend to appear in the bibliographies and databases of their disciplines, making this
research material cumbersome to locate, and reducing the opportunities to share insights,
methods, and tools across traditional disciplinary lines. As computational science
curricula and programs are developed, they are beginning to identify their own
disciplinary structure and common threads, and it was felt that a tool such as the one
proposed could be of enormous value in that effort.
The original SDSC bibliography covered not only refereed journals, but also invited
talks and meeting abstracts. Although the bibliography (and associated database) is no
longer maintained, the data on which the bibliography was based is still being collected
at SDSC by Rachel Chrisman, who as manager of operations is in charge of maintaining
SDSC’s user database10. This is significant, since many authors do not explicitly
acknowledge SDSC in the text of their publications, so that it is difficult to identify this
research except through researchers’ requests for new time allocations.
10 Time allocations are requested via written proposals that are supposed to contain references to research
accomplished under prior SDSC time-allocations. Some proposals represent "block grants" of time that
are more difficult to associate with individual pieces of research. Proposals are reviewed by the NPACI
resource allocation committee; they are received and reviewed in paper, and copies of both proposals and
reviews are archived by a commercial archiving firm, under contract to SDSC.
- page 18 - Anna Gold - 7/4/12
One interviewee argued that the task of tracking SDSC publications is a separate task
from that of tracking SDSC-funded or –supported research; SDSC publications can be
searched and browsed on the SDSC web site. On the other hand, the most recent
published bibliography incorporated a list of SDSC staff publications and presentations
as an appendix following its listing of scientific publications by SDSC-supported
A related suggestion concerning potential information systems that would benefit SDSC
was to develop a single access point for locating books, RFP’s, projects, laboratories,
research groups, and other information sources and "objects" in the areas of
supercomputing, networking, and various areas of computational science. The purpose
of such a system would be to link researchers together in ways that, due to the expansion
of SDSC activities, are becoming harder via informal and word-of-mouth practices.
PERSONAL EXPERTISE AND ASSISTANCE: Most interviewees acknowledged that
the current information environment invites a "self-service" approach to finding
information. One interviewee said that a personal service that might be helpful would be
to have relevant information "pushed" via journal routing. One said that regular training
on library systems and information strategies might be useful "for administrative staff"
such as administrative assistants. These would be more effective than merely sending
reminder e-mails to all staff about new or existing services. Presumably, generic
information tasks might be delegated to such staff if it was known that they were versed
in library procedures and resources. Another interviewee echoed this perception, that
while it is desirable for a scientist to conduct their own specialized searches, it would be
preferable to have someone on whom one could call for general reference, such as
biographical information needed for a meeting. Knowing the name and phone number of
someone they could call at the UCSD Libraries when they needed informed assistance
would be valuable in itself (a gap that might be simply filled by distributing a Libraries
business card to all SDSC staff!).
- page 19 - Anna Gold - 7/4/12
Still another interviewee pointed out that SDSC has never had a full-time professional
librarian on staff, but rather relied on half-time professionals assisted by trained student
assistants. Further, SDSC has never had a digital librarian (that is, one involved with
digital library developments) on staff, despite SDSC's important role in a variety of NSF-
funded and other digital library research and development projects. Without knowing
what such a person might be able to contribute to the organization, current staff may have
difficulty judging whether the skills of a dedicated full-time professional are needed.
Both computer centers like SDSC and libraries like UCSD's are experiencing difficulty
filling positions: SDSC has several dozen openings for programmers that they can't fill;
while UCSD Libraries are experiencing high turnover and difficulty recruiting due to pay
scales that are depressed relative to the booming high tech economy in San Diego. As a
result, both organizations are keenly aware of the value added to organizations by staff, a
value that cannot be replaced by money, bricks and mortar, or machines. One
interviewee stated that it would be most valuable if a person could be identified to help
bridge the digital library efforts of the two organizations (SDSC and UCSD Libraries).
PHYSICAL SPACE: The interviewees were asked to think about what, if anything,
would be beneficial about setting up a physical space identified with library and
information services. Those who expressed the view that this would not be likely to
benefit SDSC were also asked to imagine what they would see as the best use of such a
space if it were to be established. The responses ranged widely: several stated
unequivocally that such a space - whether used as a conventional library space or as a
more informal reading room - was not important or useful. These interviewees felt there
was more value in such a space a few years ago, but that with virtually all of the research
material they need now available either as low-cost desk copies or as web-based
resources easily accessible from their offices, they saw this as a service that is no longer
needed. If such a space existed, one argued it should be created as a part of SDSC
organizational culture, and not be a space in any sense "owned" by UCSD Libraries.
Another warned against the danger of setting up any space whose identity and purpose is
not clearly embraced and understood. In the absence of this understanding, a library
- page 20 - Anna Gold - 7/4/12
space would run the risk of being reabsorbed into generic office space (particularly under
present conditions where space is at a premium).
Given the scenario that such a space existed, the general sense was that it would be most
welcome if it housed fairly generic reference tools and small reference print collections
directly related to SDSC: a collection of SDSC print publications; some general news
resources; dictionaries, maps; perhaps an organized preprint service, or display of
preprints that might pique attention. The strongest case for establishing a physical library
might be as a physical marker for the intent to achieve the goals of all libraries: whether
digitally or otherwise. An interviewee pointed out that the development of digital
libraries serves to underscore the still powerful role of social interactions within a shared
knowledge environment. These enrich research and build community, and a physical
library identified with the work of collecting, sharing, and using knowledge objects has
an important place in any active research organization11.
DIGITAL LIBRARY SCENARIOS: In the next part of the unstructured interviews, the
respondents were read a list describing a number of digital library modules and
capabilities, and asked to say whether these sounded as though they would be useful
within the SDSC environment. Each of these modules was part of an actual integrated
web-based digital library system developed and implemented at the CERN Library12.
After answering, the interviewees were shown the live web interface for the CERN
Library site, so that they could better visualize how similar services developed for SDSC
11This need has been met at UCSD in a variety of ways by different research units and
academic departments. For example, the Center for Astrophysics and Space Sciences (CASS)
has provided a small library of core journals and books for its staff in an attractive, central
setting adjacent to laboratories and to a lounge area (the Marlar Library,
http://casswww.ucsd.edu/marlar/). The Physics Department has a small reading room
accessed by key, used to store selected core journals on site. The Center for Magnetic
Recording Research serves both local and remote corporate clientele through its small
collections and active reference service in its library, with the support of two half-time
permanent library staff.
The CERN Library system interface can be viewed at: http://weblib.cern.ch/ ; the CERN
Document System interface can be seen at: http://cds.cern.ch/ ; a set of talks about the
modules of both systems can be found at: http://documents.cern.ch/conf/cern/ .
- page 21 - Anna Gold - 7/4/12
The modules described were as follows:
A. “WebLib II”. A web interface to search and browse multimedia collection
of working documents (e-prints), books, periodicals, photos, videos
(including webcasts via RealVideo), press cuttings, all related to the work
at CERN. (see Figure 2)
B. “Conversion Server”. A file conversion service to convert different file
formats on demand.
C. “Scanning Service”: Interface where users can arrange for paper
documents to be scanned and converted. Recharge service.
D. “Electronic Document Submission”. A web interface to capture metadata
and transfer documents from author to document server. Capable of
handling many document types, version control, distribution to limited
lists; approval routing, etc. Document types include locally produced
preprints, open documents, notes, theses, photos, videos, exhibition
objects; administrative documents; “external” preprints, photos, books,
articles in proceedings, etc. Includes adding “scientific notes” to articles
or preprints. (see Figure 1)
E. “Setlink” link manager: tool to manage storing of web addresses and
extraction or concatenation of document “parts”, and do on the fly format
F. Agenda Maker/Conference manager: tool to manage web agendas of
meetings and conferences. Include document conversion, handling
transparencies, minutes, etc.
G. Distributed search (targeted “spider”) across heterogeneous
<computational science> web servers.
H. “My Library” interface supporting “My Search” profile storing; “Your
Shelf” of items from CERN library; loan account management;
- page 22 - Anna Gold - 7/4/12
FIGURE 1: CERN DOCUMENT “SUBMIT” INTERFACE
- page 23 - Anna Gold - 7/4/12
FIGURE 2: CERN WEBLIB INTERFACE
- page 24 - Anna Gold - 7/4/12
This interviewees’ response to this part of the unstructured interviews was consistently
enthusiastic about two modules in particular: the WebLib interface (module A) and the
Electronic Document Submission interface (module D). One commented that services
like these seemed to represent “a new level” of library service. Other comments were
that these would be “helpful and interesting;” “a handy thing;” “most useful; exactly what
we had hoped to create”; “very high priority;” “SDSC has this need;” “this would be very
valuable.” Some expressed high enthusiasm, but some implementation concerns, among
them ensuring that incentives existed to deposit materials centrally and / or systems
existed that could automatically gather materials from distributed sources. Several
mentioned the coincidence of these services with previous efforts or existing systems
without the full flexibility and functionality evident in the CERN systems. For example,
there has long been interest at SDSC in creating a searchable interface to media clips,
images, and visualization files; there are analogous systems in place to manage the
deposit of data for archiving, associated with the DICE data curation projects. SDSC’s
Docushare document sharing system provides some of the functionality of the
“Electronic Document Submission” module, but only for the internal organization
(intranet) and not as the infrastructure to support an archive or library (whether internal or
external). Another observed that a document system like this might work better in a
research environment like CERN’s where, despite a wide range of disciplinary interests,
there was also a clear convergence on the core research area of high- energy physics.
Regarding the other modules, opinions varied much more. The Setlink manager was not
well understood (either by the interviewer or interviewees!), contributing to the
somewhat neutral response. Most felt that the Conversion Server might be quite helpful
to other campus departments, could be useful at SDSC, but was not vital given the
convergence of file formats and the prevalence of technical knowledge sufficient to
manage conversions as needed. A similar response was received for the Scanning
Service: the sense was that this would be “pretty useful” but probably only on occasion,
- page 25 - Anna Gold - 7/4/12
reflecting in part the tendency of SDSC staff to deal primarily with documents already in
The response to the Agenda Maker / Conference Planner was more skeptical, in part
because it was felt commercial off the shelf software exists to help with such tasks
(SDSC already uses an electronic calendaring tool called Meeting Maker); and in part
because it was hard to envision how this module would interface with (or be used to
populate) an archive or library. If such a module existed and it made conferencing easier
at SDSC, “it might be useful.”
Module G offered a spider or web search capability, but focused on a pre-selected set of
known web servers whose content was thought to be of particular utility and interest at
SDSC. Although interesting, several interviewees doubted that it would be used with
confidence by staff, given the difficulty of identifying a common set of web sites and
topics equally valuable to all of SDSC’s diverse scientific, administrative, and
computational interests. One suggested that the capability would be more useful if each
group at SDSC had the capability of configuring it for their own interests.
Finally, the customizable “My Library” interface elicited the least enthusiasm from the
interviewees, some of whom described it as a “frill,” or “trivial,” though at least two
others thought it might be useful. Again, in the absence of an existing SDSC digital
library, it may be harder to envision how a user might want to customize the interface for
their personal use.
4.5 RESEARCH COLLABORATION: UNSTRUCTURED INTERVIEWS
METHOD: Interviewees in both the structured and unstructured interviews were asked
to suggest whether there were research projects at SDSC with the potential for assisting
them with their information needs. Following up on their suggestions, separate
interviews were conducted with four researchers at SDSC who were engaged in the
research efforts that had been mentioned, including work of the Data-Intensive
- page 26 - Anna Gold - 7/4/12
Computing Environments (DICE) research group; and a number of scientific
“workbench” environments building large or small integrated research portals where
data, tools, and analysis could be worked with interactively by researchers and students.
DICE: This group associated with NPACI is lead by Reagan Moore of SDSC. Their
work has taken a lead nationally in efforts to provide a common framework of analysis
across disparate data structures. Their work has also been distinguished by its
programmatic integration: using XML and advanced computer architecture, Moore's
group has developed tools to curate and provide access, navigation, display, and
archiving for extremely large collections of data.
WORKBENCHES: Other SDSC research efforts with some relevance were the
development of discipline- or task-specific portals and electronic “workbenches”
providing both information and a platform for interactivity. Two separate "workbench"
projects now in place at SDSC include the Biology Workbench, brought from NCSA to
SDSC by Shankar Subramaniam; and the Sociology Workbench, developed by Ilya
Zaslavsky (now at SDSC), while he was at San Diego State University.
The Biology Workbench developers are in discussions with the DICE group regarding
the potential for generalizing the tool using XML, and also creating a Chemistry
Workbench. The Protein Data Bank (PDB), Michael Gribskov's work on linking
biological data with Medline literature, a Biodiversity portal, and a possible new initiative
to develop an earthquake engineering portal, are some examples. One important impact
of these and similar efforts is a kind of deep restructuring of research literature, which it
seems may have fundamental effects on how literature is used in various contexts
(disciplinary, research, instruction at different levels). This in turn has important impacts
on the structures entrenched in many decades of library practices (e.g. the differentiation
between monographs and articles, the reliance on book indexes, tables of contents,
abstracts, and header information to add structure to the literature, catalog records to
provide indicators of relevance and appropriate audience, the need to locate reviews or
- page 27 - Anna Gold - 7/4/12
personally browse materials in place in order to establish the potential relevance of
literature to specific problems or tasks, etc.).
The Sociology Workbench was developed at San Diego State University (SDSU) under
the direction of Ilya Zaslavsky, a specialist in Geographic Information Systems (GIS).
The tool was written mostly in C and Java by undergraduates at SDSU, and is “a very
simple tool” that allows people to analyze any survey data online. Although the flagship
data used in the tool is the General Social Survey data, actually data can reside in any
location accessible by FTP: the tool can grab just the data for the variables in which the
researcher is interested. Under consideration now is a generalization of the tool would
require setting up procedures for upkeep, developing a different architecture to support
uniform query and extraction; it would probably depend on XML encoding of the data (a
standard being developed by the Data Documentation Initiative or DDI).
The workbench architecture not only reflects, but also supports an even deeper
transformation of research, as noted by Biology Workbench developer Shankar
Subramaniam. In the workbench architecture, data that are inherently complex and
heterogeneous can be brought together without imposing familiar categories. "The new
approach is much more Baconian: to let the data and its patterns begin to dictate the way
in which Nature will be carved at the joints. New categories will be the result of an
inductive process. I would like to see 'induction' of this sort running like a daemon in the
background on all the biological databases. Then we can take the sequences, maps, gels,
and the myriad representations of our knowledge and produce a common universe of
SIO WORKBENCH? Another suggestion was that a UCSD digital library development
should leverage the campus strength as a world leader in oceanography and biology.
Since both SDSC and UCSD Libraries have an interest in developing a digital library
Online, 1999, http://www.npaci.edu/online/v3.22/shankar.html, " Complexity, Richness and
Challenge: Shankar Subramaniam Joins UCSD/SDSC".
- page 28 - Anna Gold - 7/4/12
program relating to the Scripps Institution of Oceanography, this in turn suggests that a
workbench-like project in this area might be a promising area of collaboration.
In discussions with four SDSC researchers deeply involved in different digital library
projects, it was very clear that from their perspective, they are in the process of inventing
a digital library program that encompasses the full range of library functions, from
identification and collection, to adding navigational and contextual value, to integration
and federation of resources across space and time; to persistence and archiving. To miss
the opportunity to benefit from the remarkable commonality of interest between these
two institutions would be most unfortunate.
5. Analysis of Findings:
GENERAL: The interview findings underscored that current information habits of SDSC
staff are to rely heavily on two sources of information: first, personal interactions with
colleagues at SDSC and in the field; and second, any and all sources accessible through
the web. The majority experience a high demand on their information skills, brought
about in part by the rapid changes in technology and also by the pace of scientific change
and the trend toward multidisciplinary research. They experience the information
environment as overwhelming for the limited time they have available to spend in
research and information-seeking, factors which lead them to focus their information
searching very tightly on those subject they can filter and evaluate quickly using their
own expertise or the expertise of their peers.
PREFERRED INFORMATION SERVICES: Their information preferences
corresponded closely to their information practices, emphasizing the desire for two things
above all: first, known individuals with information expertise, to whom they could
turn as they do to other colleagues on those occasions when they need expert advice and
don’t have the time to learn a new set of skills that they are unlikely to use frequently;
and second, an integrated and clearly focused digital library system where they could
easily gain access (and contribute) to a living body of documentation and source
- page 29 - Anna Gold - 7/4/12
materials relevant to the core work and impact of their organization. The latter is
considered essential to the mission of the SDSC and was the most strongly supported
recommendation among the interviewees. The primary concern expressed was that if the
implementation was to acquire and retain both authority and usefulness, even beyond the
immediate community of the Center, it would also need to be administratively robust:
that is, it would need strong policy support from SDSC managers.
A third need that was strongly expressed was for alert services that would serve two
important functions: first, such services would help them keep current with literature and
developments (including new information tools) in their areas of interest; and second,
alert services would provide some facsimile of the serendipitous experience of browsing
in a physical library. One successful alert service used by some interviewees is
HPCwire, an e-mail delivered electronic clipping service focused on high performance
computing14. The weekly headlines cover feature stories about staff changes, new
funding, and scientific breakthroughs, as well as conference announcements, commercial
developments, and job ads. However, other interviewees expressed doubt that any single
alert service could adequately serve the wide ranging interests and needs of the SDSC
scientists, programmers, and administrators.
Some interviewees felt strongly that a fourth need was for a physical space dedicated
specifically to supporting the knowledge environment at SDSC would be important to
mark the commitment of the organization to this goal, and would acquire an ongoing
value for the organization, based on the experience of other University departments and
research centers. The importance of this space was not, primarily, to house physical
collections or machines, but rather was envisioned as a space that would support the
human interactions needed to build SDSC's knowledge environment. In the age of the
digital library, this “human” or social space has remained an important element even
while the need for physical collections has diminished significantly. In many research
14 HPC Wire, "The Text-on-Demand E-zine for High Performance Computing. Subscription
information available from email@example.com; trial subscription available from
- page 30 - Anna Gold - 7/4/12
and educational organizations, the library can and does serve as a natural focus for
important informal community interactions, even while digital library collections have
begun to dominate traditional printed materials as a primary research source. Still, many
interviewees warned that the need for physical space was offset both by the insatiable
demand for adequate office space and by changes in how they and others like them now
seek information: they feel they would be very unlikely to take advantage of such a space
and warn that a space would need unambiguous definition and unwavering management
support in order to function successfully.
Fifth, there were several more limited services that many interviewees felt would be
helpful additions to the existing suite of library and information services enjoyed now.
Among these were electronic desktop delivery of print journal articles or book chapters
held by UCSD Libraries (a service called Avanti, that is currently provided to the School
of Natural Sciences and Scripps Institution of Oceanography); specifically, color
capability was mentioned as a worthwhile addition to this service because of the
importance of color particular to biologists and chemists. Another service mentioned was
some kind of regular program of training or demos at SDSC for interested SDSC staff,
to help them keep up to speed on new databases and services. Another service mentioned
was the creation of a one-stop SDSC library “portal.” There was some disagreement
about this idea, with some regarding this function as being better served either by the
existing S&E Library web page, or by the SDSC web site, or both. It is also possible that
an SDSC online library / bibliography system might serve this need.
Finally, the several interviewees - particularly those involved in research projects -
agreed that there is a clear and present need to establish ongoing research-oriented
interactions between the SDSC and UCSD Libraries, in order to further the agendas of
both organizations. Many suggestions were made; one was that UCSD Libraries
personnel could work with SDSC to provide physical space at UCSD Libraries where
SDSC research portals might be used by researchers in a research library setting.
Appropriate equipment and software would be required, but also human assistance in
- page 31 - Anna Gold - 7/4/12
advising researchers about the appropriate and effective use of these new integrated
Another suggestion was that a UCSD Libraries programmer could be identified to
collaborate with SDSC Biology Workbench developers on generalizing research
workbenches. This might ensure that SDSC developments could reflect the needs and
capabilities of the Libraries’ clientele15.
In current research about the impact of digital libraries on traditional libraries, the phrase
"hybrid" libraries often occurs, signifying the belief that students, researchers, and
libraries will continue to demand, and interact with, an information environment in which
conventional print media and "digital" media coexist. The findings of this study indicate
that the notion of what is hybrid about a digital library of the future might need to be
revised. There is a strong sense - one might also call it an expectation - among these
interviewees that nearly all the information sources they use and need will migrate to the
web. They have respect for a small subset of printed documentation, but tend to see such
items as personal (desk copies, favorite books, personal subscriptions to magazines), and
only in very restricted cases, as candidates for sharing through a library (the interviewees
mentioned atlases, dictionaries, and generic reference tools whose conventional formats
still made them more usable than their digital counterparts). Rather, the two basic
components of hybrid libraries of the future in these interviewees' minds were the digital
15Another area where SDSC staff voiced and interest in collaboration with UCSD Libraries
was in the area of interface design and usability; a specific project where UCSD Libraries
has been invited to collaborate is the interface design for the AMICO art museum images
project. UCSD Libraries had identified an art information specialist to work on this project,
but the collaboration is not active at this time. The ad hoc and isolated nature of this
collaboration has no doubt contributed to its lack of success to date, and underscores the
need for a more programmatic approach to such collaborations. A more successful
collaboration has been the Chinese-Language Digital Library, in which SDSC has
implemented its large data storage and retrieval architecture to manage and provide access
to a collection of potentially terabyte size. See Online,
http://www.npaci.edu/online/PRDLA.html for more information.
- page 32 - Anna Gold - 7/4/12
and the human. The views of these interviewees thus unconsciously echo a statement
made many years ago by Edward Feigenbaum, a paraphrase of which goes: "The library
of the future will be a place where computers and humans collaborate." These views also
resonate strongly with those of John Seely Brown of Xerox PARC, as expressed in his
1995 essay with Paul Duguid, “The Social Life of Documents.”16 In this essay, Brown
draws attention to the instrumental role documents have always played, as objects that
help people negotiate relations between each other and in their communities.
The challenge of this study therefore is to recommend actions and projects that, when
implemented, will contribute to the development of an extensible model for a hybrid
digital-human library, one in which participation in an individual research community is
supported by and represented by extensible digital systems. It is important to underscore
that the underlying recommendation of this report is that the development of digital
library services, and the digital library research collaboration, be understood as one effort
with several parts.
Specific parts of this effort would initially include the following commitments:
1. SDSC DIGITAL LIBRARY / BIBLIOGRAPHY: Establish a joint UCSD
Libraries – SDSC team to develop a proposal and implementation plan for
creating a CERN-like system for depositing and curating a collection of citations,
documents, and media objects representing both the work of SDSC staff, and the
work of researchers resulting from their use of SDSC resources, with the
expectation of extending the bibliography to encompass work of researchers at
other supercomputing and computational science centers. The CERN Library staff
have already been contacted concerning the possibility of reusing their software,
and in reply have expressed willingness to collaborate on a project to adapt their
system design to the SDSC environment.
16John Seely Brown and Paul Duguid, “The Social Life of Documents,” 1995.
http://www.parc.xerox.com/ops/members/brown/papers/sociallife.html. Also published in
- page 33 - Anna Gold - 7/4/12
2. SDSC LIAISON: Identify a UCSD Libraries staff member to serve as an on-
going liaison to the SDSC for personal information consultations, regular
demonstrations, and to take the lead in developing the SDSC Digital Library /
Bibliography. This effort would be undertaken as a short-term (approximately
one-year) project, with a matching commitment of staffing and time by the SDSC.
These individuals could share a designated office space at SDSC, with the
expectation that they would also work to define and plan for an eventual physical
space dedicated to supporting the SDSC information environment.
3. CONVERSION AND MIGRATION SERVICES: Another recommendation is
that services and systems be developed to help with the migration of significant
research content to the web.
3.1 PRINT TO DIGITAL: The expectation that important information will be
available on the web is one that can be fulfilled in many ways, and with the
exception of the irreplaceable element of human interaction, this is an expectation
that is better served by fulfilling it than by fighting it! Fulfilling this expectation
does not necessarily require comprehensive retrospective digitization, although
many publishers and some consortia are taking that approach. The UCSD service
that offers printed documentss in digital format on demand (Avanti) is an existing,
appropriate service to bridge the digital-print divide. It is recommended that
Avanti document delivery services be extended to all SDSC affiliates, to begin
Fall Quarter 2000, on a one-year trial basis. The cost of this service based on
existing business models at the Science & Engineering Library would be modest,
probably less than $2000 per year.
3.2 VIDEO IN THE DIGITAL LIBRARY: This recommendation follows from
the widespread sense that personal interactions in workshops, conferences, and
seminars, represent invaluable resources for researchers. It is recommended that
video files (analog, webcast) be collected and converted to streaming digital
Release 1.1, October 11, 1995, and First Monday, May 1996.
- page 34 - Anna Gold - 7/4/12
formats in order to provide an asynchronous opportunity for SDSC researchers to
participate as listeners or observers. The potential impact of extending this model
across the campus would make this especially interesting, in view of the many
seminars and talks and symposia taking place at UCSD, and SDSC's interest in
developing substantive research contacts with other ORU's and departments on
3.3 ONLINE MEDIATION OF PERSONAL INTERACTIONS: A different kind
of "bridge" service is one that links people digitally. A recommendation is to
provide access to the SDSC library liaison via a live web interface, with or
without video, during real-time virtual office hours. (More elaborate, and
intrusive, web-based reference systems were not viewed favorably by the several
interviewees to whom this type of interface was described.) This service might
provide impromptu personal interaction for staff who otherwise would tend not to
pick up the phone or leave their workspace in their office. A virtual library space
(chat room) might even be developed, allowing for library-like social interactions
between SDSC staff. Tables of contents for journals, and lists of newly received
preprints might be available for browsing in this space, and others browsing the
same materials could be co-present and engage in conversations.
4. RESEARCH DISCUSSIONS: It is recommended that the UCSD Libraries and
SDSC agree to initiate an ongoing program of information exchange between
applications developers, researchers, and administrators about digital library
developments. Examples of events in such a program are to:
17The opportunity to stream video presentations of seminars and lectures has been
implemented with success at the Mathematical Sciences Research Institute (MSRI); NCSA
has been at the forefront of interactive distance education using video and webcasts. The
CERN Library interface also offers organized access to research seminars and lectures
delivered with streaming media.
- page 35 - Anna Gold - 7/4/12
UCSD Libraries include SDSC digital library researchers as panelists, co-
hosts, or guests, at presentations and visits by speakers invited by UCSD
Libraries Digital Libraries Innovation Team (DLIT).
UCSD Libraries invite SDSC digital library researchers to make
presentations to UCSD Libraries administration, DLIT, or seminars to
campus faculty are also invited.
UCSD Libraries invite comment and evaluation by SDSC researchers on
the UCSD Libraries digital library program.
Administrators and digital library researchers at SDSC and UCSD
Libraries meet quarterly to inventory and assess potential areas of
collaboration, including jointly seeking outside funding for identified
projects of mutual interest.
5. RESEARCH EXCHANGE: It is recommended that SDSC and UCSD Libraries
identify opportunities for short-term staff exchanges, for the purpose of collaborating
on identified digital library projects. The initial area of collaboration recommended is
the development of the CERN-like digital library and archive. Another promising
potential area for collaboration is in workbench development, perhaps in support of
the oceanography research community at SIO (see above).
6. COMPUTATIONAL SCIENCE AT UCSD LIBRARIES: A final
recommendation is that UCSD Libraries and SDSC work to establish a functional
presence for SDSC in a mainstream library setting where faculty and students could
be exposed to the research tools being made available by SDSC researchers: a
scientific “Informatics Lab”. This functional presence might be combined with a
recent proposal that the bioinformatics graduate program be found temporary space in
the S&E Library.
- page 36 - Anna Gold - 7/4/12
(1) email survey
(2) web site developed for Boisseau group meeting
(3) structured survey questions
- page 37 - Anna Gold - 7/4/12
Information Practices and Needs at SDSC:
Please return this survey by April 15:
o electronically: firstname.lastname@example.org
o by campus mail: Anna Gold, attn. SDSC survey, S&E Library,
o in person: bring your filled-out surveys to the briefing on April
1. What kinds of information do you use at least weekly? (check all that apply)
__ personal contacts (email, in person, by phone)
__ general interest newspapers or magazines (NYT, Science, Nature)
__ printed journals that I subscribe to or that SDSC subscribes to
__ printed journals at UCSD library
__ electronic journals
__ manuals / handbooks / directories in my office
__ internal SDSC databases or publications / reports
__ conferences and proceedings
__ workshops, courses
__ books from the UCSD library
__ books that I buy or SDSC buys
__ listservs or newsgroups
__ results from Internet searches
__ known / bookmarked web sites
__ information specialists / intermediaries / librarians
__ preprints, technical reports
__ numeric data
2. What information tool(s) do you wish for that you don't currently have? (e.g.,
database, book, application...)
3. What information service(s) do you wish for that you don't currently have? (e.g.,
information delivery option, analysis service, alert service....)
4. What would be your ideal in-house database (and interface)?
5. Describe your your ideal on-site information facility / library.
6. What references (including manuals, books, journals) would you like on site at
SDSC that are not currently there?
- page 38 - Anna Gold - 7/4/12
7. Would you be willing to participate in a 30-minute informational interview about
your information use and needs? YES ___ NO ____ If yes, please give your name
and we'll contact you to set up a convenient time: _____________________________
Thank you for your participation.
- page 39 - Anna Gold - 7/4/12
Original page at: http://library.ucsd.edu/~anna/sdsc.html
Science and Engineering Library
Libraries Portal preview (functional)
Libraries Portal preview (graphical)
Computer science journals via Portal
search serials in Roger
Some serial titles:
Computing in Science & Engineering
Fortran Forum (ACM): not online
Journal of Supercomputing
Journal of Computational Physics
Journal of Computational Chemistry
Physics Today: partial contents online
NCSA's Access Newsletter
multi-database search: SearchLight
INSPEC (physics, computer science)
Biosis, Medline, PubMed (biology)
Web of Science and Current Contents (all areas of science; citation searching)
Academic Universe and Investext (business)
Technical reports & preprints: NCSTRL; LANL
Dissertations (includes UC full text from 1996) & theses
Reference: directories, dictionaries, encyclopedias, handbooks
Gale's Directory of Associations
Books in Print
Community of Science (expertise, grants)
Books: from UCSD but also from SDSU and other local SD libraries via Circuit
Request books to be delivered to IRPS (or other UCSD) Library
Activate your library card (UCSD ID) and manage your own library account (renew
- page 40 - Anna Gold - 7/4/12
SDSC INTERVIEW QUESTIONS: version May 17 2000
My first questions are about you and your role at SDSC and UCSD:
1. Where are you in the SDSC organization chart? main functional groups:
*outreach and education
*science and research
*science and teaching
*computer/networking support (experiment, teaching)
2. How long have you been affiliated with SDSC?
3. Do you have another "home" department at UCSD (in which you teach, where you conduct
research, where you serve on committees, etc.)?
(For SDSC scientists):
My next questions are about your special area of expertise:
1. What do you consider your field or fields of expertise?
2. How has your field (fields) changed in the last five years? What are some of its current
3. What disciplines are involved in your field?
4. How would you describe the differences between your discipline and how it is practiced here,
other UCSD units
similar research centers (and what are they? NCSA....)
5. What is the importance in your field of:
preprints and tech. reports
6. What/when do you publish? where are you most likely to publish? how and with whom do
you share your work and at what points in the research cycle?
7. Who are the leaders in your field (and where are they)?
8. Do you have graduate students / research assistants with whom you work?
- page 41 - Anna Gold - 7/4/12
9. What is your perspective on your discipline: its structure, subfields?
10. How do you go about learning a new area in your field? a branch of your field? another
11. How do you keep current in your field? in the literature?
12. How does the interdisciplinary nature of your field of research affect how you find
13. How often do you find yourself needing to come up to speed in a new area?
My next questions have to do with how you find and use information generally:
1. What types of published and unpublished materials do you use most often?
2. Do you use article databases?
*no: why not?
*yes: which ones? when? do you get frustrated or are you satisfied with your skills in using
3. Do you capture the results of your searches electronically?
no: why not? would this be helpful to you?
yes: how? is this helpful to you?
4. Do you ever browse? what do you browse? where? how frequently? Is this helpful? why?
5. Do you ever delegate information gathering? If so, for what types of information? to whom? is
6. Do you subscribe to listservs? which are important and valuable to you?
7. Who or what would you go to if you needed:
a phone number or email for a researcher at another institution
an article recently published in the Journal of Computational Physics
some background or recent research relating to a presentation you'd attended
dates and names associated with work that had been done here at SDSC
help troubleshooting a program you're writing
literature search to help justify a grant proposal you're writing
8. What are your major annoyances or irritants in finding or using information? could any of these
be addressed by a library/cybrary or librarian/cybrarian?
My next questions have to do with how and when you use libraries or library services
- page 42 - Anna Gold - 7/4/12
1. Do you have any library books charged out now?
2. Do you ever go physically to any of the UCSD libraries? which ones? how often?
3. Is there any UCSD library web page that you use regularly? which one(s)?
4. Is there any other (non UCSD) library or library web page that you use regularly?
5. In your career, has there been a library or library service you particularly liked to use or that
you particularly appreciated? why?
6. Would you mind telling or showing me how you would:
renew your books (if you have any checked out)
scan the table of contents for the current issue of Science magazine?
find articles published recently in a journal important in your field?
find articles written by a new colleague?
find a book review for a recent book?
borrow a book from SDSU?
My next set of questions has to do with your own or SDSC information resources and
1. Do you keep any files or lists for your own reference? (e.g. files of article reprints,
collections of books, databases of references, etc.)
2. If you could share access to them, would any of these be useful to others at SDSC?
3. Are you aware of any research projects or applications developed at SDSC that could be
applied to help improve the SDSC information environment or services? (prompt: Biological
Workbench; FACSNET; DICE/MIX)
4. Do you maintain a bookmarks file on your web browser, or another list of useful information
links? Would you be willing to send the most important to me?
5. Are there core books or classic journal articles in your field? Would you be willing to provide a
list of these?
My last set of questions have to do with library services at/for SDSC:
1. Could you place a priority (high, medium low) on each of the following:
ALERT SERVICES/KEEPING CURRENT:
- page 43 - Anna Gold - 7/4/12
A. Alert, notification service for new research in your discipline
B. Alert, notification service for issues of general science pubs like Science
C. Alert, notification service for new web sites or library services in your area
D. Pick-up and delivery at SDSC of Library-owned physical materials
E. Desktop electronic delivery of Library-owned or ILL physical articles (Avanti)
F. Personal advice / hotline to get help using Library databases or services (e.g. webline
mediated reference or “House calls” to help you get started)
G. Personal advice to provide full-service research assistance (conduct web or database
searching for you)
H. Recommendations or help managing your own information once you locate it.
ONE-STOP / FILTERED WEB ACCESS:
I. Special / custom web site for SDSC for important and useful information
BETTER “DATA-MINING” of SDSC INSTITUTIONAL MEMORY, EXISTING RESOURCES:
J. Mechanism for sharing library / research recommendations with colleagues (mini
book reviews etc.)
K. Help managing or accessing SDSC-produced publications and information
OTHER SERVICE OR PROJECT NOT MENTIONED:
WHO ELSE SHOULD I BE INTERVIEWING?