Building The European Digital Library - An Insider’s Point of View

Building The European Digital Library -
An Insider’s Point of View
Olaf D. Janssen
Olaf D. Janssen is a project manager for The European Library Office, based in the National Library of The Netherlands in
The Hague, The Netherlands
Email : Olaf.Janssen@theeuropeanlibrary.org

Abstract
In December 2004 Google announced its plans to digitise and publish millions of books from 5 prestigious Anglo-Americ-
an academic libraries by the year 2015. Initiated by French fears that Google's initiative could create a bias towards Anglo-
American language and culture, Europe quickly united to mobilise funds for the digitisation, preservation and accessibility
of European cultural heritage and the creation of a European Digital Library, including 6M digital works from libraries,
museums and archives by 2010.
Today The European Library (TEL) is a multilingual portal offering integrated access to the tens of millions of re-
sources of Europe's national libraries. It offers free federated searching and delivers digital objects - some free, some
priced.
The EU stressed that the European Digital Library should not be constructed from scratch, but build on existing initi-
atives such as TEL, because TEL has a long history of successfully implementing and using some of the vital ingredients
for the European Digital Library. These include A) internal & external collaboration and cooperative organisational net-
works, B) a technological platform based on creating, maintaining and conforming to common standards in i) data harvest-
ing and access protocols, ii) metadata and iii) collection descriptions and C) multilingual access.
The reader will 1) learn what it takes to build a pan-European Digital Library, 2) find out about the history and future
of this project and C) discover that this a win-win-win project: for its users, for its builders, and for world knowledge.

1. Introduction and History tries, supporting French President Chirac in asking
for coordination and funding from the European
In December 2004 Google announced 1 its plans to Union (EU) to create a European Digital Library.
digitise and publish online 15 million volumes This initiative also found the support of a broad
from 5 prestigious Anglo-American academic lib- coalition of 24 European national libraries.
raries by the year 2015. This bold initiative sparked Making the holdings of Europe's libraries, mu-
a wave of activities across Europe. seums and archives available online is not a trivial
In an article 2 in the French newspaper Le task. There is a wide range of different materials
Monde the President of the French national library available: books, film fragments, photographs,
expressed his concern that Google's initiative manuscripts, sheet music, speeches, sounds, etc.
could create a bias towards the English language Furthermore, what materials to select from around
and Anglo-American culture, especially for future 2.5 billion books and bound periodicals in
generations. He stressed that diversity and multi- Europe's libraries & archives and millions of hours
lingualism are basic values of the European culture of film and video in its audiovisual archives?
that need to be protected and preserved. His com- The EU recognised the vital importance to 1)
ments were widely picked up in the media, who digitise, 2) preserve and 3) open up Europe's written
immediately presented it as a 'cultural war with and audiovisual heritage on the internet. In other
Google'. words, to make it usable for European citizens, in-
Things began to speed up when the French call novators, artists and entrepreneurs for their stud-
for safeguarding the European cultural heritage ies, work or leisure, for now and for future genera-
got critical backing from the leaders of five coun- tions. In March 2006 the EU announced 3 financial,

strategical and organisational support for building erated access to the tens of millions of resources
a European Digital Library. The decision to co- (books, magazines, journals etc. - both digital and
fund the creation of a Europe-wide network of di- non-digital) of currently 23 national libraries.
gitisation centres and to address the issues of Aimed at both professional and non-professional
copyright protection was welcomed by all stake- researchers and informed citizens world-wide, it
holders with great enthusiasm. offers free searching in a vast virtual collection of
The EU proposed the following timeline: materials from all disciplines. It delivers digital ob-
• 2008 - Multilingual access to digital jects - some free, some priced.
collections of national libraries. The The current (April 2007) participants in TEL
collections must be searchable and are the CENL members of Austria, Croatia,
usable. A minimum of 2 million digit- Cyprus, Czech Republic, Denmark, Estonia, Fin-
al works (books, pictures, sound files land, France, Germany, Hungary, Italy(2x), Latvia,
etc.) should be accessible through the Lithuania, Malta, Netherlands, Poland, Portugal,
European Digital Library. Serbia, Slovakia, Slovenia, Switzerland and the UK.
The collections and catalogues of the remaining 24
• 2010 - The European Digital Library CENL members will be included at a later stage.
needs to have expanded to include Most notably, in September 2006 an 18 month
collections of a number of archives, project8 started to include the 9 remaining EU and
museums and other libraries, and pos- EFTA9 countries in TEL. Next year the latest EU
sibly publishers. A minimum of 6 mil- members Bulgaria and Romania will also be added.
lion digital works should be accessible For understanding the rest of this article, it is
through the European Digital Library. important to realise that The European Library is
In practice, this number could well be not only
much higher, if cultural institutions of a very useful technical platform (in the form of
different types and at different levels the TEL portal7) for the end-users, because
(national, regional, local) participate.
• Of the phenomenal depth and quality
of trusted deep web resources held in
the national libraries.
2. The Situation Today
• It gives easy access to native re-
2.1 CENL sources held in other countries.
As is clear from this historical introduction, na- • It enables types of collection-level
tional libraries are the forerunners where building searching which would otherwise be
the European Digital Library is concerned. This is impossible.
no surprise, because European national libraries - • It is a major contribution to research
as opposed to national museums and archives - both in making resources widely
have a long and successful history of collaboration, available and by making possible new
going back as far as 1987. In that year the Confer- connections through exploitation of a
ence of European National Librarians 4 was foun- huge virtual library collection.
ded. CENL aims to increase and reinforce the role but also
of the national libraries in Europe, in particular in a very important organisational collaborative plat-
respect of their responsibilities for maintaining the form for the participating national libraries, because
national printed heritage.
Members of CENL are the national libraries of • It provides an international showcase
the member states of the Council of Europe 5 and for their collections, products and ser-
Vatican City. The Conference currently consists of vices.
47 libraries6 from 45 European countries (Italy and • It gives them an increased exposure
Russia have two national libraries each). on the world stage with combined
political mass providing greater mar-
2.2 The European Library (TEL) keting and negotiation power.
One of the activities of CENL is the operation
and development of The European Library (TEL)
portal 7. Since 2005 it offers multilingual and fed-

• It gives libraries a feedback loop on CENL supports the idea that TEL is a model
what users are expecting on a platform and model organizational network for
European scale. building the European Digital Library. As a group,
the members of CENL own Europe's cultural
• It is a mechanism to extend collabora- published heritage; many of them according to leg-
tion. al deposit, many for the whole period of time of
• It provides feedback on user demands their nation's history.
which can prioritise institutional and From the above it will be clear that TEL - and
national digitisation activities. CENL as its founding organisation – are in an ex-
• TEL provides a cooperative frame- tremely advantageous position where building the
work for continuous development, European Digital Library is concerned. No other
sharing and innovation in metadata, organisations are mentioned so explicitly as cata-
interoperability and other technical lysts to create a common multilingual access point
standards. for Europe’s collective memory. For them
however, this does not come as a surprise, as
2.3 Current position of TEL and CENL CENL and TEL have been building and experi-
A true European Digital Library should serve all menting with some of the vital ingredients for the
types of user needs: present and future, up-to-date European Digital Library for a long time.
and historical information, science and humanities,
education, research and everyday, normal informa-
tion needs. It must comprise all types of media 3. Three Pillars for the European Digit-
from the full range of Europe's cultural heritage al Library
institutions.
Regarding the technology of a proposed
European Digital Library, it was clear from the be- As a response to the statements and recommenda-
ginning that any kind of central database will be tions of the EU, CENL had to re-think the posi-
impossible to achieve. Similar to the TEL portal, tion of TEL within the bigger framework of the
integrated multilingual access to the digitised ma- European Digital Library. This is still very much
terials of Europe's libraries, archives, museums will an ongoing process, but for now TEL sees itself as
be more realistic. The contents of the European the key player in one of the three main pillars of
Digital Library will thus grow at the same speed as the European Digital Library:
the underlying digital collections in the participat- 1) Providing online access to the digital materials.
ing institutions. This means that TEL is not directly in-
In its March 2006 announcement3 the EU volved in the two other key areas of the
stressed that the European Digital Library should European Digital Library already men-
not be constructed from scratch, but build on ex- tioned in the introduction:
isting initiatives in the European cultural heritage 2) Mass-digitisation of as many materials as
field. The EU explicitly stated that possible and
“[the European Digital Library] will build upon the 3) Long-time digital preservation of these digit-
TEL-infrastructure, currently the gateway to the catalogue ised materials.
records of collections in a number of national libraries, Before looking in more detail on how TEL is
which also gives access to a range of digitised resources of the giving users online access to cultural heritage now
participating libraries.” Furthermore, in August 2006 and in the future, first some brief words about 2)
the EU officially recommended 10 mass-digitisation and 3) digital preservation.
“[the European Digital Library] would make it possi-
ble to search Europe's distributed - that is to say, held in 3.1 Existing digital content and mass-digitisa-
different places by different organisations – digital cultural tion
heritage online. Such an access point would increase its visi- In the present TEL portal there are some rich
bility and underline common features. The access point seams of digitised materials, such as photographs,
should build on existing initiatives such as The European maps, music scores, manuscripts and posters.
Library (TEL), in which Europe’s libraries already col- However, there is an emphasis on catalogue re-
laborate…..” cords, rather than digital resources. Given the fact
that currently only 5-10% of the metadata is en-

riched by digital objects, there is a desperate need 4) The technical challenge of how to lower
for more digital content on TEL. costs for digitisation & preservation while
Part of this problem can be solved relatively maintaining a high quality.
easy, as a recent survey across 25 CENL members 5) The legal challenge of how to deal with
has shown that large numbers of digitised objects the copyright aspects in cooperation with
are readily available to be made accessible. The right holders in order to ensure coverage
National Library of Spain for instance has some 40 of protected works.
million scanned newspaper pages waiting to be
OCRed and put online. This survey also showed 3.2 Digital preservation
that many institutions are starting or are already As is clear from 3.1, most of what has been said
running their own digitisation programmes, mainly about the challenges of digitisation equally applies
focusing on special sets of materials (e.g. rare to the hurdles of long-term digital preservation.
books, manuscripts, incunabula, old newspapers). Several EU-funded projects are currently dealing
Because of this, CENL has set an overall goal with the specifics of digital preservation in the
for the coming years to digitise more content more scope of the European Digital Library. The most
quickly and to make sure that access is as complete important initiatives are:
as possible. This also includes investigating what • CASPAR - Cultural, Artistic and Sci-
means are already or could become available with- entific knowledge for Preservation,
in and across the libraries to make more efficient Access and Retrieval 12
use of existing content-rich collections and invest-
igating ways to facilitate the creation of virtual • PLANETS - Preservation and Long-
content-rich collections across (and within) the lib- term Access through NETworked
raries. Services 13
However, all these fragmented activities are • DPE – Digital Preservation Europe 14
just drops in the ocean of 2.5 billion books and
bound periodicals available in Europe's libraries.
What is really needed - and forms the current de- 4. Ingredients for Success in the
bate between European cultural heritage stake-
holders - is a coordinated large scale mass digitisa- European Digital Library
tion & preservation programme where the follow-
ing challenges are addressed : The rest of this paper will concentrate on three of
1) The economic challenge of who will pay the vital ingredients CENL and TEL are using
for the digitisation & preservation. The currently to successfully provide online access to
EU made clear 11 it will not pay for the ac- the holdings of the European national libraries and
tual digitisation work; this is left to nation- on how these same building blocks can be used
al and institutional levels. However, it will and expanded to give users access to the wider
co-fund the creation of a Europe-wide cultural heritage of libraries, museums and
network of centres of competence for digitisa- archives in the European Digital Library by 2010.
tion & preservation. These centres will
house the skills and expertise needed to 4.1 Internal & external collaboration and co-
achieve excellence for digitisation & operative organisational networks
preservation processes. They will integrate At the very core of the current success of CENL
and build on existing know-how in tech- lies its 20 year history of internal and external collab-
nology companies, universities, cultural in- oration. Since 1987 European national libraries
stitutions, and other relevant organisations. have been building a firm cooperative organisa-
2) The organisational challenges of how to tional network in which they have set aside their
create synergies and avoid duplication of differences and focused on their commonalities.
effort in cultural institutions and how to They have streamlined their international policies,
secure public-private collaboration. have experimented with improving online access
3) The challenge of content selection. What to their catalogues and collections and have de-
to choose from the thousand of kilometres veloped and shared innovations in standards for
of books, newspapers and periodicals and metadata, interoperability and access mechanisms.
millions of hours of film and video?

Participants in TEL have been able to take ad- finding ways of jointly presenting
vantage of this internal cohesion to create a com- content to the user.
bined political mass which has widened their ex- • BRICKS – Building Resources for Integ-
posure on the European and world stage and in- rated Cultural Knowledge Services 22
creased their negotiation and marketing powers. TEL is also in contact with BRICKS.
As there still is no such thing as a “CENL/TEL This project works with museums,
for museums or archives”, the EU had not much libraries and other organisations and
difficulty in finding the best starting point for their aims to maximise the impact for the
plans to build the European Digital Library. construction of a shared digital herit-
External collaboration has always been a prior- age, which nevertheless respects the
ity for CENL. Most relevant for this paper is its European cultural diversity. Its peer-
ongoing dialog with the Federation of European to-peer approach maximises the use
Publishers (FEP)15. FEP represents 25 national as- of existing resources and know-how,
sociations of book and learned journal publishers and, therefore, national investments.
in EU and EFTA member states. Other cultural
and scientific heritage associations and networks Dr. Elisabeth Niggemann, Chair of CENL,
CENL is cooperating with include eIFL16, IFLA17, says23 the following about cooperative networks in
LIBER18 and UNESCO19. the European Digital Library:
To define and reinforce its position for build- “A European Digital Library will consist of
ing the European Digital Library, in 2006 TEL highly diverse materials held by institutions with
started working together with a number of relevant different professional backgrounds and traditions
initiatives in the European cultural heritage field. in countries with different institutional structures,
Among these are responsibilities and financing. It will therefore be
• DELOS – Network of Excellence on Di- nearly impossible to organise a central, compre-
gital Libraries 20 hensive super-structure. Instead, very similar to
This network intends to conduct a TEL, a networked structure is required, allowing
joint program of activities aimed at faster and slower partners to proceed at their own
integrating and coordinating the on- speeds, while all benefit from the process.
going research activities of the major The 'hub' of the network should be a central
European teams working in Digital entry point to all the participating gateways. This
Library-related areas with the goal of network of networks must be scalable, and will rely
developing the next generation Digit- heavily on common rules, standards and proced-
al Library technologies. ures. On the other hand, it must be built with di-
The cooperation between DELOS versity and heterogeneity in mind. Partner net-
and TEL focuses primarily on the in- works need not be homogenised, but can continue
tegration of DELOS-provided func- to be organised in many ways. They will create
tionality into the existing TEL portal among themselves a network of subnetworks, with
and ensures that duplication of activ- nodes and substructures, that reflect the diverse
ity does not take place. needs of the different user communities, media
• MICHAEL – Multilingual Inventory of types, institution types, and eventually also reflect
Cultural Heritage in Europe 21 legal frameworks.
This project focuses on the integra- Since the Europe of the future will be larger
tion of national initiatives in digitisa- than it is now, it is important that all European
tion of the cultural heritage and inter- countries are taken into consideration from the
operability between national cultural very beginning, not only today's EU/EFTA states.
portals to promote access to digital The European Digital Library should also – from
contents from museums, libraries and the very beginning - try to build bridges to those
archives. TEL is collaborating with global or regional networks outside Europe that
MICHAEL to find ways of building provide additional resources for Europe's citizens
on each others knowledge using the and researchers.” [end quote]
expertise within the networks and

4.2 Technological platform based on common by the TEL portal ensures interoperability between
standards disparate collections.
The current technological platform for The The advantages of having OAI harvested
European Library – the TEL portal - is based on cross-collection metadata in one central database
creating, maintaining and conforming to common are numerous:
standards in • Searching in one Central Index deliv-
A. Data harvesting and access protocols, er the results to the end-user much
B. Metadata and faster than it can be done via distrib-
C. Collection descriptions. uted searching in remote collections.
Because of the network of the 47 CENL mem-
ber libraries there is already a large-scale, working • In user studies on federated searching
and proven implementation of these standards at in digital library portals it is shown
an international level; one more reason why the again and again that end-users are
EU is looking at TEL/CENL as one of the cata- 'Google-minded': they don't under-
lysts for the European Digital Library. stand the fact that they first have to
An explanation of the TEL technical architec- select the collections they want to
ture is given by Van Veen and Oldroyd in the Feb- search in before they can make a
ruary 2004 issue of D-Lib Magazine 24. Although query. They expect to search in one
the portal has been considerably revised and exten- single Google-like blob. Having all
ded over the last 3 years, the basic principles have TEL metadata in a Central Index
remained the same. Some relevant points from this solves this problem.
paper have been extracted here. • The TEL portal can do on-the-fly
background searches in the Central
A: Common standards in data harvesting and access proto- Index and suggest to the end-user po-
cols tentially interesting objects or related
One of the sources of bibliographic records for collections he did not actively search
the TEL portal is a central repository of metadata for. Furthermore , ranking, merging
obtained by harvesting national library collections and de-duplication of search results
via the Open Archives Initiative Protocol for becomes possible.
Metadata Harvesting 25 (OAI-PMH). The cata- • Personalization services, multilingual
logue records are harvested and indexed to form information access and other func-
the so-called TEL Central Index. This is compar- tionalities are very hard to implement,
able to how for instance Google works. if possible at all, on a distributed
For small static collections (i.e. few new research model. They rely on having the
cords are added and records are not changed of- metadata stored centrally.
ten) harvesting via FTP is also an option, although
this methods is currently used for only a handful
• All in all: for most patrons of the
TEL portal OAI brings a much better
of TEL collections.
user experience.
Not all ±100 million records available in TEL
today are hold in the Central Index. As explained The OAI protocol, its purpose and advantages
in the Van Veen & Oldroyd paper 24 the TEL are known in many, but not all European national
portal is a distributed search engine : remote col- libraries. Mostly because it is a fairly recent tech-
lections that are not (yet) OAI harvestable (thus nology, it does not have a long history, unlike e.g.
whose metadata is not available in the Central In- Z39.50, that has been around in libraries for dec-
dex) are searchable on-the-fly via the SRU pro- ades. Currently an OAI promotion campaign is
tocol 26. In case the remote target supports SRU running aimed at those libraries that are not yet
this is a direct and seamless process. However, if fully aware of what OAI can mean for their collec-
the remote system only ‘speaks’ Z39.50, commu- tions. Other libraries are working towards OAI-
nication has to go via a Z39.50/SRU gateway, compatibility of (part of) their collections. This in-
which acts as a protocol, query language and char- cludes:
acter set translator. Although somewhat less
powerful than OAI, the support of these protocols

a) Building new or adapting existing infra- leads to the proposal of a European Metadata Re-
structure/hardware to support OAI access gistry.
to collections In an article 31 in ERCIM News 66, László
b) Making metadata OAI-compliant: i.e. Kovács, András Micsik and Jill Cousins discuss the
cross-walking and converting to TEL Ap- idea of a European Metadata Registry:
plication Profile 27 “The need for this type of joint European re-
OAI is widely regarded by most European cul- gistry is obvious: National Libraries in Europe ap-
tural heritage institutions as the way forward to ply different legacy metadata schemas, The
promote digital content to portals and search en- European Library uses the TEL Application Pro-
gines. Thus OAI-PMH and its inherent central file to ensure interoperability when performing a
storage of metadata will inevitably become corner- search across libraries and collections.
stones of the European Digital Library. The European Metadata Registry would
provide a set of services:
B: Common standards in metadata • It would describe different metadata
The TEL portal relies on a common metadata ap- schemas and/or application profiles,
plication profile 27 which is based on the Dublin as well as the aims, target audiences,
Core Metadata Element Set 28 and the DC-Library application circumstances and scope
Application Profile 29. This means that all currently of the schemas.
±100 million objects (in 250+ collections 30) avail-
• It would represent internal semantic
able in TEL share a common metadata format.
structure, the hidden model of
Libraries that want to offer collections via the
schemas. Model descriptions aim both
TEL portal have to make sure the records 1) are
to understand and document the
formatted in XML and 2) comply with the TEL
terms hierarchy. Because the metadata
Application Profile for Objects 27. For certain re-
schemas of partners are based on dif-
cord formats, e.g. the MARC-family, this means
ferent background semantical models,
cross-walking or conversion of metadata before it
model mappings are non-trivial and
can be included in the portal. To assist libraries in
require further scientific investiga-
this process, automatic conversion tools are avail-
tions.
able in the portal architecture.
Understanding that the TEL portal will evolve • It would register tools and/or on-line
over time and that new functionality will be re- services available for mappings, infer-
quired as new collections of different types of me- encing, translations, versioning and
dia become available, means that the TEL Applica- access.
tion Profile is not static. TEL has a mechanism to • Finally, it would register semantic
enable the metadata model to evolve over time in connections and relations between
a controlled way : the TEL metadata registry. different schemas, thereby fostering
This registry is a database that records all the the reuse of profiles, terms, elements
metadata activity associated with TEL. It therefore and encoding schemas.
contains not only all the elements that are part of With these services the EMR will support the
TEL Application Profile, but also elements that production of relevant crosswalks between legacy
have been proposed by different partners but have metadata schemas and/or The European Library
not yet been accepted for the application profile, schema…[ ]. The registry can grow organically on
and those elements that have been rejected. the basis of functional granularity and bilateral
Whilst the registry is an essential part of TEL, mappings. The scope and scalability of the registry
it can also serve the much wider – and for the are also under study, since the number of re-
European Digital Library essential - purpose of gistered schemas and mappings cannot be pre-
sharing metadata definitions outside the TEL ser- dicted.
vice. It can form a mechanism for the European We intend the EMR to be a standardized, yet
national libraries to share metadata functionality flexible and user-friendly tool. It will administer
with other European cultural heritage institutions, European metadata from different European cul-
such as museums and archives. Extending this idea tural heritage communities including libraries, mu-

seums and archives, and its aim is to ensure trans- must be able to deal with all the European lan-
parency, access and interoperability.” [end quote] guages with their different character sets. This is
also a vital requirement from a more abstract point
C: Common standards in collection descriptions of view: multilingualism and thus cultural diversity
Collection descriptions allow libraries to provide are core values of European culture. As the
information about the existence and availability of Europe of the future will be larger than it is now it
their collections, not only to direct end-users, but is also very important to address from the very be-
also e.g. to the big search engines like Google. De- ginning all European countries, not only those that
scribing collections is vitally important for feder- are today's EU Member States. Whilst English is
ated searching. The TEL portal provides access to effectively the lingua franca of Europe, construct-
many heterogeneous collections, therefore, in or- ing a single-language portal would seriously ob-
der to perform more focussed searching, standard- struct many potential clients from using the portal
ised collection descriptions have to be available for in a way they are entitled to. Despite the fact that
each of the currently 250+ collections. Collection many are catching up quickly, not all Europeans
descriptions enable users to narrow down the list speak English, especially older populations in most
of collections, which at the second step become EU New Member States 35.
targets of a (more precise or focused) search. The current TEL portal gives users multilin-
Collection descriptions in TEL are 1) XML gual access to collections and objects in the fol-
files that 2) comply with the TEL Application Pro- lowing ways:
file for Collection Descriptions 32. They also com- • The full user interface has been trans-
ply with the NISO Collection Description Spe- lated into 20 European languages; this
cification 33 and the related NISO Metasearch Ini- number will grow to ± 30 by 2008.
tiative Collection Description Schema 34. TEL Although this set will still not cover
member libraries are required to provide rich, all official languages of Europe, it will
complete, high quality descriptions - to foster mul- include the native or secondary lan-
tilingual access (see 4.3) at least in their native languages of over 99% of (potential)
guage and in English, but preferably in all TEL clients.
European languages. They must make sure that
the descriptions are written with end-users in • As stated in 4.2.C, TEL partner librar-
mind, so for instance give meaningful, ‘non-librari- ies are required to describe their col-
an’ names to their collections. lections in at least their native lan-
Similar to the TEL metadata element set, col- guage and English. To increase multi-
lection descriptions in TEL are not static, but can lingual access, TEL members are
evolve over time where needed; not only to reflect presently working on a coordinated
new functionality or new types of access, but also effort to translate all 250+ descrip-
to cater for a wider variety of organisations. Most tions into all 20 currently supported
of what has been said under 4.2.B about metadata languages. The resulting approx. 5000
thus equally applies to collection descriptions. Mu- collection descriptions will help the
seums and archives have their own specific tradi- vast majority of users in better finding
tions and requirements for describing their collec- what they are looking for.
tions, backgrounds to be respected in the • The TEL system is (nearly) Unicode
European Digital Library. However, sharing and compliant. In the past diacritics have
re-using TEL collection descriptions outside the caused some problems, particularly in
scope of TEL will have huge potential benefits for the display of results, but now UTF-8
interoperability of European cultural heritage. MI- has been implemented across all TEL
CHAEL21 is currently gathering collection de- infrastructural components and (most
scriptions from a range of non-national-library in- of) the metadata.
stitutions. The present set of multilingual tools is still a
far cry from how a fully multilingual European Di-
4.3 Multilingual access gital Library needs to function. Obviously, the
From the beginning the TEL portal has been set ideal scenario will need to be developed step-by-
up as a scalable system. This not only means tech- step, probably by different teams of different dis-
nical scalability, but also functional scalability: it ciplines in different countries. A first step is cur-

rently being taken by investigating cross-language facilitate the forthcoming digitisation process and
access to bibliographic subject data (e.g. in con- online accessibility of works protected by copy-
trolled vocabularies). A selection of cross-language right and related rights. Voluntary contractual ar-
approaches to subject data, including MSAC36, rangements between representatives of cultural in-
MACS and its child CrissCross37 will be tested in stitutions and rights holders are proposed as part
the TEL portal. of the Commission Recommendation on the Digit-
“The MACS project 38 aims to provide multi- isation and Online Accessibility of Cultural Materi-
lingual subject access to library catalogues. It en- al and Digital Preservation 10.
ables users to simultaneously search a number of 2) Marketing & Communication - A big and ongo-
catalogues in the languages of their choice. These ing effort will be needed to market, promote and
include the most used first and second languages communicate the assets of the European Digital
in European: English, French, German and Span- Library to its (potential) end-users. To a certain ex-
ish. This multilingual search is made possible tent this should be done via traditional offline mar-
thanks to the equivalence links created between keting & communication methods (newspapers, ra-
the four indexing languages: SWD/RSWK (for dio, TV, trade press, conferences etc.), but it
German), RAMEAU (for French), LCSH (for should largely focus on online methods (search en-
English) and the Spanish subject headings lan- gine optimisation 42, linking programmes, targeted
guage.” emailing etc. ). Not only should this involve a
Work is also being done in the field of name central effort on behalf of all participating institu-
authority control. The results of projects such as tions, but far more should it concentrate on local,
LEAF39, VIAF40 and ONESAC41 are used to more bespoke marketing & communication ef-
design and test a name authority control tool, en- forts, so promoting the European Digital Library
abling the TEL portal to locate and use in a query from within the partners countries and institutions,
any known variations of person, corporate body or in the native language, according to local best-
geographical names. practice methods and targeted at the specific local
The ideal European Digital Library would audiences.
need to offer true multilingual search & retrieval,
particularly where full-text corpora are concerned.
This means searching for a particular phrase in (for 5. And Finally .. ..the European Digital
instance) English and finding not only documents Library vs. Google
that contain that exact English phrase, but also
(probably non-English) documents containing the The entire idea of building a European Digital
translated equivalent of the original query. It Library was sparked off by Google’s announce-
would also need to include non-Latin European ment to digitise millions of books from 4 Americ-
languages, such as Cyrillic. Evidently, for full-texts an and 1 English academic library. By various me-
digital objects, both the object itself and its de- dia this was presented as a 'cultural war with
scriptive metadata would need to be queried to Google'. So far the European cultural institutions
give the most satisfying and useful search results. have never had this notion of a ‘competition’.
For non-text digital items (maps, photos, sounds They have always been pragmatic: they stress their
etc.) one would need to particularly rely on own strengths, realise their weaknesses and are
metadata, as content-based retrieval for such ob- aware of how private enterprises impact their
jects is - similar to true multilingual full-text search worlds.
& retrieval - still in its very early stages. Libraries, museums and archives have spent
thousands of man years giving structure to data
4.4 Other ingredients for success and separate the wheat from the chafe. Pre-quali-
So far 3 major components have been discussed fied structured (meta)data has a very important
that ensure the current success of TEL and will role to play where accurate and reliable search &
contribute significantly to the eventual success of retrieval in trusted sources are concerned. This is
the European Digital Library. To finish this sec- not something Google can (yet) handle with its
tion, two further ingredients needed for a fruitful mass market search engine approach. So for the
end-result are briefly mentioned: time being heritage institutions seem to beat
1) IPR issues - issues concerning intellectual Google at its own mission to ‘organize the world's
property rights need to be solved in order to fully

information’ …… at least that of the cultural Eventually the real truth will be somewhere in
world. between, with well-working public-private partner-
For centuries Europe’s collective memory has ships, with useful, easy-to-find, accurate and trust-
been collected, maintained and safeguarded by worthy information and – most importantly - with
specialised, dedicated, sustainable, respected, trus- happy customers. Obviously, building the
ted, non-commercial, public institutions. Cultural European Digital Library - already called “one of
heritage preservation has never been - and, for the the greatest digital construction efforts ever under-
sake of global learning, should never become - a taken - is really is a win-win-win project: for its
solely commercial enterprise. Why should this be users, for its builders, and for world knowledge.
different for the digital age?

1
Endnotes and References
E

Google Checks Out Library Books - the libraries of Harvard, Stanford, the University of Michigan, the University of Oxford,
and The New York Public Library join with Google to digitally scan library books and make them searchable online, 14 De-
cember 2004 http://www.google.com/press/pressrel/print_library.html
2
Jeanneney, Jean-Noël, “Quand Google défie l’Europe”, Le Monde, 24 January 2005
http://www.bnf.fr/pages/dernmin/pdf/articles/lemonde_2401.pdf
3
European Commission steps up efforts to put Europe’s memory on the Web via a “European Digital Library”, 2 March 2006
http://europa.eu/rapid/pressReleasesAction.do?reference=IP/06/253
4
CENL – Conference of European National Librarians http://www.cenl.org
5
The Council of Europe's Member States http://www.coe.int/T/E/Com/About_Coe/Member_states/default.asp
6
The National Libraries of Europe http://libraries.theeuropeanlibrary.org
7
TEL – The European Library http://www.theeuropeanlibrary.org
8
EDLproject http://www.edlproject.eu
9
EFTA - European Free Trade Association http://www.efta.int
10
European Commission Recommendation on the Digitisation and Online Accessibility of Cultural Material and Digital
Preservation, 24 August 2006 http://ec.europa.eu/information_society/newsroom/cf/document.cfm?
action=display&doc_id=160
11
The European Digital Library: Frequently Asked Questions, 2 March 2006 http://europa.eu/rapid/pressReleasesAction.do?
reference=MEMO/06/102
12
CASPAR - Cultural, Artistic and Scientific knowledge for Preservation, Access and Retrieval http://www.casparpreserves.eu
13
PLANETS - Preservation and Long-term Access through NETworked Services
http://www.planets-project.eu
14
DPE – Digital Preservation Europe http://www.digitalpreservationeurope.eu
15
FEP – Federation of European Publishers http://www.fep-fee.be
16
eIFL - Electronic Information for Libraries http://www.eifl.net
17
IFLA - International Federation of Library Associations and Institutions http://www.ifla.org
18
LIBER - Ligue des Bibliothèques Européennes de Recherche http://www.kb.dk/liber
19
UNESCO - United Nations Educational, Scientific and Cultural Organization http://www.unesco.org
20
DELOS - Network of Excellence on Digital Libraries http://www.delos.info
21
MICHAEL - Multilingual Inventory of Cultural Heritage in Europe http://www.michael-culture.org
22
BRICKS - Building Resources for Integrated Cultural Knowledge Services http://www.brickscommunity.org

Niggemann, Elisabeth, “The European Digital Library - A Project of the Conference of European National Librarians”,
23

ERCIM News, 66, July 2006 http://www.ercim.org/publication/Ercim_News/enw66/niggemann.html

Van Veen, Theo and Bill Oldroyd, “Search and Retrieval in The European Library: A New Approach”,
24

D-Lib Magazine, 10(2), February 2004 http://www.dlib.org/dlib/february04/vanveen/02vanveen.html

25
Open Archives Initiative Protocol for Metadata Harvesting - v.2.0 http://www.openarchives.org/OAI/openarchivespro-
tocol.html
26
SRU - Search and Retrieve via URL http://www.loc.gov/standards/sru
27
The European Library Application Profile for Objects – Version 1.5 http://www.theeuropeanlibrary.org/metadatahand-
book/tel_ap.html
28
Dublin Core Metadata Element Set - Version 1.1 http://dublincore.org/documents/dces
29
DC-Library Application Profile (DC-Lib) http://dublincore.org/documents/library-application-profile
30
Overview of collections in The European Library http://www.theeuropeanlibrary.org/collections
31
Kovács, László, András Micsik and Jill Cousins, “Towards The European Metadata Registry”, ERCIM News, 66, July 2006
http://www.ercim.org/publication/Ercim_News/enw66/kovacs.html
32
The European Library Application Profile for Collections Descriptions – (old version 1.5) http://www.theeuropeanlib-
rary.org/metadatahandbook/tel_ap_cld.html
33
NISO Collection Description Specification http://www.niso.org/standards/resources/Z39-91-DSFTU.pdf
34
NISO Metasearch Initiative Collection Description Schema http://www.ukoln.ac.uk/metadata/nisomi/cd-schema
35
New Member States of the EU : Estonia, Latvia, Lithuania, Cyprus, Malta, Poland, Hungary, Slovakia, Czech Republic and
Slovenia
36
MSAC - Multilingual Subject Access to Library Catalogues http://www.nkp.cz/_en/pages/page.php3?page=fond_sub-
jectauthorities_eng_pozn1.htm
37
CrissCross http://www.ddb.de/eng/wir/projekte/crisscross.htm
3
38

MACS – Multilingual Access to Subjects https://macs.vub.ac.be
3
39

LEAF - Linking and Exploring Authority Files http://web.archive.org/web/*/http://www.leaf-eu.org/
40
VIAF - Virtual International Authority File http://viaf.org
41
ONESAC - ONE Shared Authority Control http://www.portia.dk/websites/onesac.htm
42
SEO - Search Engine Optimisation http://en.wikipedia.org/wiki/Search_engine_optimisation

Suggested further reading:
ERCIM News - Special theme: European Digital Library, ERCIM News 66, July 2006
http://www.ercim.org/publication/Ercim_News/enw66

Building The European Digital Library - An Insider’s Point of View

Recommended

Recommended

More Related Content

What's hot

What's hot (11)

Viewers also liked

Viewers also liked (20)

Similar to Building The European Digital Library - An Insider’s Point of View

Similar to Building The European Digital Library - An Insider’s Point of View (20)

More from Olaf Janssen

More from Olaf Janssen (20)

Recently uploaded

Recently uploaded (20)

Building The European Digital Library - An Insider’s Point of View