Nof S.Y. Springer Handbook of Automation

Подождите немного. Документ загружается.

1285

Library Autom

72. Library Automation

Michael Kaplan

Library automation has a rich history of 130 years

of development, from the standardization of card

catalogs to the creation of the machine-readable

cataloging (MARC) communications format and

bibliographic utilities. Beginning in the early 1980s

university libraries and library automation ven-

dors pioneered the ﬁrst integrated library systems

(ILS). The digital era, characterized by the prolifer-

ation of content in electronic format, brought with

it the development of services for casual users as

well as scholarly researchers — services such as

OpenURL linking and metasearching and library

staff tools such as electronic resource manage-

ment systems. Libraries are now reacting to user

demands for quick, easy, and effective discovery

and delivery such as those they have grown ac-

customed to through the use of Google and other

Internet heavyweights, by developing a new series

of Library 2.0-based discovery-to-delivery (D2D)

applications. These newest offerings deliver an

up-to-date user experience, allowing libraries to

retain their back-ofﬁce systems (e.g., acquisitions,

cataloging, circulation) and add to or replace them

as needed. In the process libraries can leverage

Web 2.0 services to interconnect systems from

72.1 In the Beginning:

Book Catalogs and Card Catalogs............1285

72.2 Development of the MARC Format

and Online Bibliographic Utilities ..........1286

72.2.1 Integrated Library Systems ............1287

72.2.2 Integrated Library Systems:

The Second Generation .................1289

72.2.3 The Nonroman World:

Unicode Comes to Libraries............1290

72.3 OpenURL Linking and the Rise

of Link Resolvers ..................................1290

72.3.1 Metasearching .............................1291

72.3.2 The Digital Revolution

and Digital Repositories ................1293

72.3.3 Electronic Resource Management ...1293

72.3.4 From OPAC to Next-Generation

Discovery to Delivery.....................1294

72.4 Future Challenges.................................1296

72.5 Further Reading ...................................1296

72.5.1 General Overview .........................1296

72.5.2 Speciﬁc Topics..............................1296

References ..................................................1297

different vendors and thereby ensure a gradual

transition toward a new automation platform.

72.1 In the Beginning: Book Catalogs and Card Catalogs

In many ways one can date the beginning of post-

industrial era library automation to the development

of the library catalog card and the associated card

catalog drawer [72.1, 2]. The nature of the origi-

nal library catalog card can be gleaned from this

patent description of the so-called continuous li-

brary catalog card that had evolved to take ad-

vantage of early computer-area technology by the

1970s [72.1]:

a continuous web for library catalogue cards hav-

ing a plurality of slit lines longitudinally spaced

7.5cmapart, each slit line extending 12.5cmtrans-

versely between edge carrier portions of the form

such that upon removal of the carrier portions

outwardly of the slit lines a plurality of standard

7.5cm×12.5cmcatalogue cards are provided.Lon-

gitudinally extending lines of uniquely shaped feed

holes or perforations are provided in the carrier

Part G 72

1286 Part G Infrastructure and Service Automation

Fig. 72.1 Example of a handwritten catalog card (courtesy

of University of Pennsylvania [72.3])

portions of the form which permit printing of the

cards by means of existing high speed printers of

United States manufacture despite the cards be-

ing dimensioned in the metric system and the feed

of the printers being dimensioned in the English

system.

Card catalogs and card catalog drawers are well de-

scribed in this patent that pertains to an update to the

basic card catalog drawer [72.2]:

an apparatus for keeping a stack of catalogue cards

in neat order to be used, for example, in a library. If

necessary, a librarian can replace a damaged cata-

logue card with a new one by just pressing a button

at the bottom of the drawer to enable a compression

spring to eject the metal rod passing through a stack

of catalogue cards. Consequently, the librarian can

freely rearrange the catalogue cards.

Though these devices continued to evolve, the orig-

inals, when standardized at the insistence of Melvil

Fig. 72.2 Example of OCLC-produced catalog (shelﬂist)

card (courtesy of OCLC [72.4])

Dewey in 1877, led eventually to the demise of large

book catalogs, which, despite their longevity, were at

best unwieldy and difﬁcult to update [72.5]. As we will

see later, the goal here was one of increased produc-

tivity and cost reduction, not that different from the

mission espoused by late 20th century computer net-

works.

The development and adoption of the standard cat-

alog card presaged the popular Library of Congress

printed card set program that, together with a few

commercial imitators, was the unchallenged means of

universalizing a distributed cataloging model until the

mid 1970s. The printed card set program, in turn, was

succeeded by computerized card sets from the Ohio

College Library Center (OCLC), later renamed the

OCLC Online Computer Library Center, Inc.; the Re-

search Libraries Group (RLG); and other bibliographic

utilities. This was largely due to the advantage the utili-

ties had of delivering entire production runs of card sets

that could be both customized in format and at the same

time delivered with card packs already presorted and

alphabetized.

72.2 Development of the MARC Format and Online Bibliographic Utilities

The success of the bibliographic utilities coincided

with several other developments that together led to to-

day’s library automation industry. These developments

were the MARC communication format, the develop-

ment and evolution of university homegrown and then

vendor-based library systems, and ﬁnally the growth

of the networking capabilities that we now know as

the Internet and the World Wide Web. In fact, with-

out the development of the MARC format, library

automation would not have been possible. Taken as

a whole, these milestones bookend 130years of library

automation.

Two extraordinary individuals, Henriette Avram and

Frederick Kilgour, were almost single-handedly re-

sponsible for making possible library automation as

we know it today. During her tenure at the Library

of Congress, Avram oversaw the development of the

MARC format; it emerged as a pilot program in 1968,

Part G 72.2

Library Automation 72.2 Development of the MARC Format and Online Bibliographic Utilities 1287

became a US national standard in 1971, and an inter-

national standard in 1973 [72.6]. The MARC standard

bore a direct connection to the development of OCLC

and the other bibliographic utilities that emerged dur-

ing the 1970s. RLG, the Washington Library Network

(WLN), and the University of Toronto Library Au-

tomation System (UTLAS) all sprang from the same

common ground.

Kilgour, who had served earlier at Harvard and Yale

Universities, moved to Ohio in 1967 to establish OCLC

as an online shared cataloging system [72.7]. OCLC

now has a dominant, worldwide presence, serving over

57000 libraries in 112 countries with a database that

comprises in excess of 168million records (including

articles) and 1.75billion holdings.

From the beginning the goals of library automation

were twofold: to reduce the extremely labor-intensive

nature of the profession while increasing the level of

standardization across the bibliographic landscape. It

was, of course, obvious that increasing standardization

should lead to reduced labor costs. It is less clear that

actions on the bibliographic “production room ﬂoor”

made that possible. For many years these goals were

a core part of the OCLC mission – the most successful

modern (and practically sole surviving) bibliographic

utility. Founded in 1967, OCLC is a [72.8]

nonproﬁt, membership, computer library service

and research organization dedicated to the public

purposes of furthering access to the world’s infor-

mation and reducing information costs.

In the early years of library automation, libraries em-

phasized cost reduction as a primary rationale for

automation.

In the early 1970s a number of so-called biblio-

graphic utilities arose to provide comprehensive and

cooperative access to a database comprised of descrip-

tive bibliographic (or cataloging) data – now commonly

known as metadata. One by one these bibliographic

utilities have vanished. In fact, in 2007 OCLC absorbed

RLG and migrated its bibliographic database and bib-

liographic holdings into OCLC’s own WorldCat, thus

leaving OCLC as the last representative of the utility

model that arose in the 1970s.

A hallmark of library automation systems is their

evolution from distinct, separate modules to large,

integrated systems. On the level of the actual biblio-

graphic data this evolution was characterized by the

migration and merger of disparate ﬁles that served dif-

ferent purposes (e.g., acquisitions, cataloging, authority

control, circulation, etc.) into a single bibliographic

master ﬁle. The concept of such a master ﬁle is pred-

icated on avoidance of duplicative data entry; that

is, data should enter the system once and then be

repurposed however needed. A truly continuous, es-

calating value chain of information, from publisher

data to library bibliographic data, still does not ex-

ist. Libraries, especially national libraries such as the

Library of Congress, routinely begin their metadata

creation routines either at the keyboard or with subop-

timal data (often from third-party subscription agents).

(The Library of Congress also makes use of an elec-

tronic cataloging-in-publication program.) The online

information exchange (ONIX) standard presages the ul-

timate extinction of the pure MARC communication

standard as extensible markup language (XML)-based

formats become dominant. Both the hardware/software

and the data strains become ever-more intertwined as

bibliographic automation becomes a subset of the larger

information universe, adopting standards that transcend

libraries.

72.2.1 Integrated Library Systems

The story of modern library automation began with

a few pioneering individuals and libraries in the mid

to late 1960s [72.9–13]. In addition to Avram and Kil-

gour, oneof the more notable was Herman Fussler at the

University of Chicago, whose efforts were supported

by the National Science Foundation. Following the ini-

tial experiments, the university began to develop a new

system, one of whose central concepts was in fact the

bibliographic master ﬁle [72.9].

The University of Chicago was not alone. Stan-

ford University (with bibliographic automation of large

library operations using time sharing (BALLOTS)),

the Washington Library Network, and perhaps most

importantly for large academic and public libraries,

Northwestern University (NOTIS) were all actively

investigating and developing systems. NOTIS was em-

blematic of systems developed through the mid 1980s.

NOTIS and others (e.g., Hebrew University’s auto-

mated library expandable program (Aleph 100) that

became the seed of Ex Libris and the Virginia Tech

library system (VTLS)) were originally developed in

university settings. In the 1980s NOTIS was commer-

cialized and later sold to Ameritech in the heyday

of AT&T’s breakup into a series of Baby Bells. The

NOTIS management team eventually moved on and

started Endeavor Information Systems, which was sold

to Elsevier Science and then in turn to the Ex Libris

Group in 2006.

Part G 72.2

1288 Part G Infrastructure and Service Automation

Other systems, aimed at both public and academic

libraries, made their appearance during the 1970s and

1980s. Computer Library Services Inc. (CLSI), Data

Research Associates (DRA), Dynix, GEAC, Innovative

Interfaces Inc. (III), and Sirsi were some of the bet-

ter known. Today only Ex Libris, III, and SirsiDynix

(merged in 2005) survive as major players in the inte-

grated library system arena.

The Harvard University Library epitomizes the vari-

ous stages of library automation on a grand scale. Under

the leadership of the fabled Richard DeGennaro, then

Associate University Librarian for Systems Develop-

ment, Harvard University’s Widener Library keyed and

published its manual shelﬂist in 60 volumes between

1965 and 1979 [72.14]. At the same time the university

was experimenting with both circulation and acquisi-

tions applications, the latter with the amusing moniker,

computer-assisted ordering system (CAOS), later re-

named the computer-aided processing system (CAPS).

In 1975 Harvard also started to make use of the rela-

tively young OCLC system. As with other institutions,

Harvard initially viewed OCLC as a means to more

efﬁciently generate catalog cards [72.15].

In 1983 Harvard University decided to obtain the

NOTIS source code from Northwestern University to

unify and coordinate collection development across the

100 libraries that constituted the vast and decentral-

ized Harvard University Library system. The Harvard

system, HOLLIS, served originally as an acquisitions

subsystem. Meantime, the archive tapes of OCLC trans-

actions were being published in microﬁche format as

the distributable union catalog (DUC), for the ﬁrst time

providing distributed access to a portion of the Union

Catalog – a subset of the records created in OCLC.

It was not until 1987 that the catalog master ﬁle was

loaded into HOLLIS. In 1988 the HOLLIS OPAC (on-

line public access catalog) debuted, eliminating the

need for the DUC [72.15].

It was, in fact, precisely the combination of the

MARC format, bibliographic utilities, and the emer-

gence of local (integrated) library systems that together

formed the basis for the library information architecture

of the mid to late 1980s. Exploiting the advantages pre-

sented by these building blocks, the decade from 1985

to 1995 witnessed rapid adoption and expansion in the

ﬁeld of library automation, characterized by maturing

systems and increasing experience in networking. Most

academic and many public libraries had an integrated

library system (ILS) in place by 1990. While the early

ILS systems developed module by module, ILS systems

by this time were truly integrated. Data could ﬁnally

be repurposed and reused as it made its way through

the bibliographic lifecycle from the acquisitions mod-

ule to the cataloging module to the circulation module.

Some systems, it is true, required overnight batch jobs

to transfer data from acquisitions to cataloging, but true

integration was becoming more and more the norm.

By the early 1990s libraries were entering the age

of content. Telnet and Gopher clients made possible

the online presence of abstracting and indexing services

(A&I services). Preprint databases arose and libraries

began to mount these as adjuncts to the catalog proper.

DOS-based Telnet clients gave way to Windows-based

Telnet clients. Later in the 1990s, Windows-based

clientsinturngavewaytowebbrowsers.

All thistime, theunderlying bibliographic databases

were largely predicated on current acquisitions. The

key to an all encompassing bibliographic experience

lay in retrospective conversion (Recon) as proven by

Harvard University Library’s fundamental commitment

to Recon. Between 1992 and 1996 Harvard University

added millions of bibliographic records in a concerted

effort to eliminate the need for its thousands of cata-

log card drawers and provide its users with complete

online access to its rich collections. Oxford University

and others soon followed. While it would be incorrect to

say that Recon is a product of a bygone era, most major

libraries do indeed manage the overwhelming propor-

tion of their collection metadata online. This has proven

to be the precursor to the massive digitization projects

underwritten by Google, Yahoo, and Microsoft, all of

which depend to a large degree on the underlying bib-

liographic metadata, much of which was consolidated

during the Recon era.

In the early years of library automation, library

systems and library system vendors moved from time-

sharing, as was evident in the case of BALLOTS,to

large mainframe systems. (Time-sharing involves many

users making use of time slots on a sharedmachine.) Li-

brary system vendors normally allied themselves with

a given hardware provider and specialized in speciﬁc

operating system environments. The advent of the Inter-

net, andmore especiallythe World WideWeb, fueledby

the growth of systems based on Unix (and later Linux)

and relational database technology (most notably, Or-

acle) – all combined with the seemingly ubiquitous

personal computer – gave rise to the second genera-

tion in library automation, beginning around 1995. As

noted, the most visible manifestation of that to the li-

brary public was the use of browsers to access the

catalogs. This revolution was followed within 5years

by the rapid and inexorable rise of e-Content. The

Part G 72.2

Library Automation 72.2 Development of the MARC Format and Online Bibliographic Utilities 1289

most successful library automation vendors, it is now

clear, are those that possessed a clear and compelling

vision for combining library automation with deliv-

ery of content. The most successful of those vendors

also espouse an e-Content philosophy that is publisher

neutral.

72.2.2 Integrated Library Systems:

The Second Generation

Libraries, often at the direction of their governing bod-

ies, made a generational decision during this period

to migrate from mainframe to client–server environ-

ments. Endeavor Information Systems, founded with

the goal of creating technology systems for aca-

demic, special, and research libraries, was ﬁrst out of

the starting blocks and had its most successful pe-

riod for about a decade following its incorporation in

1994 [72.16]. It was no accident, given its roots in

NOTIS and its familiarity with NOTIS software and

customers, that Endeavor quickly captured the over-

whelming share of the NOTIS customer base that was

looking to move to a new system. This was partic-

ularly true in the period immediately preceding the

pivotal year 2000, when Y2K loomed large. In plan-

Fig. 72.3 Endeavor (now Ex Libris) WebVoyáge display

ning for the shift to the year 2000 library computer

systems faced the same challenge all other legacy com-

puter faced, namely, reworkingcomputer codeto handle

calendar dates once the new millennium had begun.

So many libraries, both large and small, had to re-

place computer systems in the very late 1990s that

1998 and 1999 were banner years in the library com-

puter marketplace, followed by a short-term downturn

in 2000.

Ex Libris Ltd., a relative latecomer to the North

American library market, was founded in Israel in 1983,

originally as ALEPH Yissum, and commissioned to de-

velop a state-of-the-art library automation system for

the Hebrew University in Jerusalem. Incorporated in

1986 as Ex Libris, a sales and marketing company

formed to market Aleph, Ex Libris broke into the Eu-

ropean market in 1988 with a sale to CSIC, a Spanish

network of 80 public research organization libraries

afﬁliated with the Spanish Ministry of Science and

Technology. This sale demonstrated an early hallmark

of Aleph, namely deep interest in the consortial envi-

ronment.

At roughly the same time as Endeavor was releasing

its web-based client–server system, Voyager, Ex Lib-

ris released the fourth version of the Aleph system,

Part G 72.2

1290 Part G Infrastructure and Service Automation

Fig. 72.4 Catalog record from University of Iowa InfoHawk catalog, with parallel roman-alphabet and Chinese ﬁelds

(courtesy of University of Iowa)

Aleph 500. Based on what was to become the industry

standard architecture, Alephfeatured a multi-tier client–

server architecture, support for a relational database,

and – a tribute to its origins and world outlook – sup-

port for the unicode character set. Starting in 1996, with

Aleph 500 as its new ILS stock-in-trade, Ex Libris cre-

ated a US subsidiary and decided to make a major push

into the North American market. At the same time –

and crucially for its success in the emerging e-Content

market – Ex Libris made the strategic commitment to

invest heavily in research and development with the

stated goal of creating a uniﬁed family of library prod-

ucts that would meet the requirements of the evolving

digital library world.

72.2.3 The Nonroman World:

Unicode Comes to Libraries

If there is a single accepted analysis of the state of li-

braries and library automation in the 21st century, it is

the fact the library world is but one part of the much

larger information universe and that library standards,

e.g., the MARC format and the American standard code

for information interchange (ASCII) character set, need

to be considered in light of the larger universe of infor-

mation standards, such as XML (ONIX) and unicode.

Those libraries and library vendors that have the most

robust understanding of this fundamental shift have

been most adept at long-term survival.

72.3 OpenURL Linking and the Rise of Link Resolvers

The shift from ASCII to unicode and from MARC to

XML heralded an entire salvo of digital products that

introduced a new concept of library automation – the

ILS was no longer the be all and end all of library au-

tomation – and was marked by Ex Libris’s release of

SFX. SFX was developed in conjunction with Herbert

Van de Sompel (presently a research scientist at the

Los Alamos National Laboratory) and the Universiteit

Ghent in response to what is known as the appropriate

copy problem [72.17]:

Part G 72.3

Library Automation 72.3 OpenURL Linking and the Rise of Link Resolvers 1291

Link

server

Full text

e-Print

ILL/

Doc Del

Reference

manager

Digital

collections

A&I

Portal

OPAC

Fig. 72.5 SFX and OpenURL linking

(after [72.4])

There has been an explosive growth in the number

of scholarly journals available in electronic form

over the internet. As e-Journal systems move past

the pains of initial implementation, designers have

begun to explore the power of the new environment

and to add functionality impossible in the world

of paper-based journals. Probably the single most

important such development has been reference

linking, the ability to link automatically from the

references in one paper to the referred-to articles.

Led by Oren Beit-Arie (presently Ex Libris’s chief

strategy ofﬁcer), SFX (which stands for special ef-

fects) was not only a commercial success but also

became generically associated with the standard known

as OpenURL linking. In much the same way as Xerox

is synonymous with photocopying, so SFX has become

synonymous with context-sensitive linking. Released in

2000, SFX now has an installed base of over 1500 li-

braries and research centers worldwide

Librarians, especially systems librarians, are well

known as staunch proponents of standards. Thus it was

an extremely astute move when Ex Libris took the

initiative to have the National Information Standards

Organization (NISO) adopt the OpenURL as a NISO

standard.

The OpenURL framework for context-sensitive ser-

vices was actually approved on a fast-track basis as

ANSI/NISO Z39.88 in 2005 [72.18]:

The OpenURL framework standard deﬁnes an

architecture for creating OpenURL framework ap-

plications. An OpenURL framework application is

a networked service environment, in which pack-

ages of information are transported over a network.

These packages have a description of a referenced

resource at their core, and they are transported with

the intent of obtaining context-sensitive services

pertaining to the referenced resource. To enable

the recipients of these packages to deliver such

context-sensitive services, each package describes

the referenced resource itself, the network context

in which the resource is referenced, and the context

in which the service request takes place.

72.3.1 Metasearching

Ex Libris quickly followed its success with SFX by ex-

panding into the area of metasearching. Also known

as federated searching, metasearching was another re-

sponse to the proliferation of online journals and online

content. MetaLib (and competing products) were the

initial bibliographic answer to the problem posed by

commercial search engines such as Yahoo or Google,

namely that much content had come quickly to live

outside the realm of the library catalog. Metasearch

enables users – power and nonpower users, whether

the general public or undergraduates, graduate students,

post-doctoral students, and faculty – to retrieve the

Part G 72.3

1292 Part G Infrastructure and Service Automation

Fig. 72.6 Boston College MetaQuest, with quick sets

most comprehensive set of results in a single search,

including traditional bibliographic citations, abstract-

ing and indexing service information, and the actual

fulltext whenever it is available. Before long, numer-

ous offerings of both OpenURL link resolvers and

metasearch applications were available in the biblio-

graphic arena.

By 2005 many institutions had made their primary

bibliographic search gatewaynot the library catalog, but

their metasearch application. Boston College is a good

example. Its library catalog, Quest, is but one of the

numerous targets searchable through its metasearch ap-

plication, MetaQuest.

By now libraries and their users had bought whole-

heartedly into the combination of metasearching and

linking – it is the combination of these two approaches

for dealing with the e-Content universe that makes

them together vastly more powerful than either alone.

Troubles loomed, though, in the absence of standards

for metasearching. NISO once again stepped forward

and created the NISO Metasearch Initiative. NISO de-

scribed the challenge as [72.19]:

Metasearch, parallel search, federated search,

broadcast search, cross-database search, search

portal are a familiar part of the information com-

munity’s vocabulary. They speak to the need for

search and retrieval to span multiple databases,

sources, platforms, protocols, and vendors at one

time. Metasearch services rely on a variety of ap-

proaches to search and retrieval including open

standards (such as NISO’s Z39.50), proprietary

API’s, and screen scraping. However, the absence

of widely supported standards, best practices, and

tools makes the metasearch environment less efﬁ-

cient for the system provider, the content provider,

and ultimately the end-user.

Standards are invaluable, but in the library world

standards – aimed at providing consistency in search-

ing – are all the more important. As knowledge seekers

move from one environment to the next with the change

of a uniform resource locator (URL), they also expect

consistent behavior across these various informationen-

vironments. It is particularly encouraging, therefore, to

see how the public responds. An article in the Octo-

ber 7, 2005 Harvard Crimson shows the opinion of

a student at one of the world’s most prestigious univer-

sities [72.20]:

Of the many things to change at Harvard over

the summer...there is one new development that

students may not have noticed but that come term-

paper time will make a world of difference. That

change is the switch from Harvard University Li-

brary’s (HUL) e-Resources website to the new

e-Research @ Harvard Libraries website. HUL has

done a tremendous job with the neweasy-to-use site,

Part G 72.3

Library Automation 72.3 OpenURL Linking and the Rise of Link Resolvers 1293

and we applaud them for dreamingup such a helpful

resource.

The new e-Research page puts a high-tech user in-

terface on top of Harvard’s vast array of online

subscriptions, making resources more readily ac-

cessible – whether users know the exact journal

they are looking for or are just poking around for

sources. But the best thing about the new e-Research

page is the plethora of features it provides that

take full advantage of computers to make research

easier ... within seconds you can ﬁnd what you

are looking for, either in fulltext online or a Find

it @ Harvard button that tells you exactly where in

Harvard’s huge library system to look.

Within two short paragraphs two of the major tools

of 21st century librarianship and e-Scholarship emerge:

metasearching (e-Research) and OpenURL linking

(Find It @ Harvard). The library catalog (HOLLIS) is

mentioned, too, but it is but one amid a vast array of

electronic resources.Looming over all this is thespecter

of Google, whose seemingly endless series of innova-

tions and content-related initiatives is in turn pushing

the frontiers of automation within the library systems

world itself.

72.3.2 The Digital Revolution

and Digital Repositories

Three other areas of institutional bibliographic automa-

tion were under discussion and development between

2002 and 2007: digital repositories, electronic resource

management systems, and new models of resource dis-

covery and delivery.

As with early developments in library automation,

one of the ﬁrst efforts to manage digital objects, espe-

cially textual objects, began in auniversity environment.

In 2000 the Massachusetts Institute of Technology

Libraries and Hewlett-Packard embarked on a joint re-

search project that we now know as DSpace – an early

digital asset management (DAM) system.

DSpace introduced the concept of an institutional

repository (IR) [72.21],

a robust, software platform to digitally store...

collections and valuable research data, which had

previously existed only in hard copies...anarchiv-

ing system that stores digitalrepresentations of ana-

log artifacts, text, photos, audio and ﬁlms...capable

of permanently storing data in a non-proprietary

format, so researchers can access its contents for

decades to come.

DSpace is an open-source software toolkit. Au-

tomation vendors realized this was a fertile ﬁeld for

institutions not inclined to go the toolkit approach. Var-

ious products have emerged; notable among them are

CONTENTdm, developed at the University of Wash-

ington and now owned by OCLC; and Ex Libris’s

DigiTool. An open-source newcomer to the ﬁeld is

Fedora, which will likely appeal to the same group of li-

braries interested in DSpace [72.22]. All these products

are designed for deposit, search, and retrieval. While

they have maintenance aspects about them, they are not

true preservation systems.

Libraries and archives, as recognized custodians

of our intellectual and cultural heritage, have long

understood the value of preservation and ensuring lon-

gitudinal access to the materials for which they have

accepted custodianship. As more and more materials

are digitized, particularly locally important cultural her-

itage materials, or are born digital, these institutions

have begun to accept responsibility for preserving these

materials. Recently (2007) the National Library of New

Zealand leapt to the forefront of digital preservation

by developing a National Digital Heritage Archive

(NDHA) program to ensure the ongoing collection,

preservation, and accessibility of its digital heritage col-

lections. This multiyear program, begun in conjunction

with Endeavor Information Systems and continued by

Ex Libris, aims to preserve digital objects for nothing

less than the life of the custodial bodies [72.23].

72.3.3 Electronic Resource Management

On the serial side of the bibliographic house, the

e-Journal evolution of the 1990s turned rapidly into the

e-Journal avalanche of the 21st century. Libraries and

information centers that subscribed to thousands or tens

of thousands of print journals found themselves facing

a very different journal publication model when pre-

sented with e-Journals. Not only have libraries had to

contend with entirely new models – print only, print

with free e-Journal, e-Journal bundles or aggregations,

often in multiple ﬂavors from competing publishers

and aggregators – but also these e-Journals came with

profoundly different legal restrictions on use of the

e-Content. Libraries found themselves swimming up-

stream against a rushing tide of e-Content. A number

of institutions (among them Johns Hopkins University,

Massachusetts Institute of Technology, and the Univer-

Part G 72.3

1294 Part G Infrastructure and Service Automation

sity Libraries of Notre Dame) developed homegrown

systems to contend with the morass; most others relied

on spreadsheets and quantities of paper.

Into this fray stepped the Digital Library Founda-

tion (DLF), which convened a small group of e-Journal

and e-Content specialists to identify the data elements

required to build an electronic resources management

(ERM) system. This study, developed in conjunction

with a group of vendors and issued in 2004 as Elec-

tronic Resource Management: Report of the DLF ERM

Initiative (DLF ERMI), was one of the few times since

the development of the MARC standard that a standard

predated full-blown system development [72.24].

Systems developers adopted the DLF ERMI initia-

tive to greater or lesser degrees as dictated by their

business interests and commitment to standards. Un-

like the OpenURL link resolvers and metasearch tools,

electronic resource management tools addressed an au-

dience that was primarily staff, harking back to the

original goal of library automation: to make librarians

more productive.

Innovative Interfaces Inc. was the ﬁrst vendor to re-

lease an ERM product. Serials Solutions, one of the

newer players, whose origins lay in providing catalog

copy for e-Journals and A-to-Z lists for e-Journals, has

the 360Resource Manager offering.

72.3.4 From OPAC to Next-Generation

Discovery to Delivery

A deeply held tenet of librarians is that they have

a responsibility to provide suitably vetted, authorita-

tive information to their users. In the years since the

rise of Google as an Internet phenomenon, librarians

have shown enormous angst in the reliance, particularly

among the generation that has come of age since the

dawn of the Internet, on whatever they ﬁnd in the ﬁrst

page of hitsfrom any ofthe mostpopular Internet search

engines. Critical analysis and the intellectual value of

the resulting hits are less important to the average user

than the immediacy of an online hit, especially if it oc-

curs onthe ﬁrstresults page. Libraries continue to spend

in the aggregate hundreds of millions of dollars on con-

tent, both traditional and electronic, yet have failed to

retain even their own tuition-paying clientele.

OCLC, to its credit, confronted the issue head on

with a report it issued in 2005 [72.25]. In the report

OCLC demonstrates quite conclusively that ease of

searching and speedy results, plus immediate access to

content, are the most highly prized attributes users de-

sire in search engines. Unfortunately, the report shows

that libraries and library systems rank very low on this

scale. On the other hand, when it comes to a question of

authoritative, reliable resources, the report shows that

libraries are clear winners. The question has serious im-

plications for the future of library services. How should

libraries and library system vendors confront this co-

nundrum?

The debate was framed in part by a series of lec-

tures thatDale Flecker, associate director of the Harvard

University Library for Planning and Systems, gave in

2005 on: OPACS and our changing environment: ob-

servations, hopes, and fears. The challenge he framed

was the evolution of the local research environment that

comprises multiple local collections and catalogs. Har-

vardUniversity, for example, had (and stillhas) separate

catalogs for visual materials, geographic information

systems, archival collections, social science datasets,

a library OPAC, and numerous small databases. Li-

censed external services have proliferated: in 2005

Harvard had more than 175 search platforms on the

Harvard UniversityLibrary portal – all in addition to In-

ternet engines, online bookstores, and so forth. Flecker

expressed hope that OPACs would evolve to enable

greater integration with the larger information envi-

ronment. To achieve this, OPACs (or their eventual

replacements) would have to cope with both the Inter-

net and the explosion of digital information that was

already generating tremendous research and innovation

in search technology. Speed, relevance (ranking), and

the ability to deal with very large results sets would be

the key to success – or failure [72.26].

Once again, the usual series of library automation

vendors stepped up. In truth, they were already work-

ing on their products but this time they were joined by

yet another set of players in the library world. North

Carolina State University adopted Endeca [72.27], best

known as a data industry platform, to replace its library

OPAC: “Endeca’s unique information access platform

helps people ﬁnd, analyze, and understand information

in ways never before possible” [72.27]. As the NCSU

authors Antelman et al. noted in the abstract to their

2006 paper [72.28],

Library catalogs have represented stagnant tech-

nology for close to twenty years. Moving toward

a next-generation catalog, North Carolina State

University (NCSU) Libraries purchased Endeca’s

Information Access Platform to give its users

relevance-ranked keyword search results and to

leverage the rich metadata trapped in the MARC

record to enhance collection browsing. This pa-

Part G 72.3