![[INET'99]](/web/20081006173754im_/http://www.isoc.org/isoc/conferences/inet/99/proceedings/inet99sm.gif)
![[ Up ]](/web/20081006173754im_/http://www.isoc.org/isoc/conferences/inet/99/proceedings/up.gif)
![[Prev]](/web/20081006173754im_/http://www.isoc.org/isoc/conferences/inet/99/proceedings/prev.gif)
![[Next]](/web/20081006173754im_/http://www.isoc.org/isoc/conferences/inet/99/proceedings/next.gif)
Kim H. VELTMAN <k.veltman@mmi.unimaas.nl>
Maastricht McLuhan Institute
Netherlands
This paper outlines some of the major initiatives under way to provide multimedia access to cultural heritage around the world (e.g., UNESCO, G7, European Commission) and focuses on two fundamental problems of the Internet: the need for authority lists and the need to include historical dimensions of knowledge. Thousands of projects are concerned with digitizing the contents of museums, galleries, and libraries. The new networks of institutions are beginning to provide thesauri and authority lists (often called ontologies in the United States), metadata, and other standards for exchanging cultural heritage information. Terms also change with time. The concept of digital reference rooms offers a method to address these two problems. Without standards in a networked environment, there will be no interoperability and the results will remain isolated. Digital reference rooms offer a new key to interoperability and provide ways of understanding historical and cultural dimensions of knowledge. This will transform the role of search engines and fundamentally change the scope of the Internet as we now know it. This paper builds on themes introduced in an earlier presentation at INET'96 in Kuala Lumpur. There some of the implications for education were explored. Here the larger implications for cultural heritage, knowledge organization, and the future role of the Internet will be addressed.
Part one of the paper will review major initiatives to provide networked access to cultural heritage. These include the work of the Consortium for Interchange of Museum Information, the Dublin Core of the Ohio Computer Library Center, bodies such as the International Council of Museums, the Museum Documentation Association, the Canadian Heritage Information Network, the Marburg Archive, the Getty Trust, UNESCO's Memory of the World Project, and G7's pilot project 5: Multimedia Access to World Cultural Heritage. It will report on the European Commission's projects in cultural heritage (e.g., MOSAIC, Aquarelle) and its new MEDICI (Multimedia for Education and employment through Integrated Cultural Initiatives) Framework, which pursues the initiatives of the memorandum of understanding for multimedia access to Europe's cultural heritage. MEDICI has four main action lines: intermuseum thematic virtual multimedia exhibitions; education, tourism, and best-practice lists. In addition, it is initiating a European Network of Cultural Heritage Centers of Excellence, which will be spearheaded by the new Maastricht McLuhan Institute (MMI), linked with a secretariat in Milan and other major centers such as the Scuola Normale Superiore in Pisa.
These initiatives are introducing common standards to permit interoperability among distributed collections, which have two fundamental implications for the Internet: One entails the organization of knowledge itself while the other involves a historical and cultural approach to knowledge. The question of how best to collect knowledge has traditionally led to a spectrum of responses. At one extreme was the notion of a central depository which gathers everything in one place: a dream that inspired the library of Alexandria and the British Library but does not work because no single building is big enough for the whole of knowledge. At the other extreme is the model of a completely distributed system, much like the Internet at present, which does not work because there are no common standards for names, subjects, places, and chronologies, that is, no authority lists to permit interoperability.
Interim solutions to this challenge include metadata (e.g., Dublin Core) and thesauri. Part two of the paper suggests that centralized, digital reference rooms offer a new answer to problems of distributed, networked knowledge. Reference rooms are the traditional equivalents of search engines and structuring tools. Indeed, the reference rooms of libraries have served as civilization's cumulative memory concerning search and structure methods through classification systems, dictionaries, encyclopedias, book catalogs, citation indexes, abstracts, and reviews. Digital reference rooms thus offer keys to more comprehensive tools.
There is a second fundamental problem. At present the Internet is focused primarily on the latest news as if it were a kind of electronic newspaper. The creators of search engines assume that isolated keywords in the form of natural language and translated into different languages will solve all our problems. Natural language does not account for variant spellings. More important, it does not account for historical and cultural shifts in the meanings of terms. The word science, for instance, is readily translated into German as Naturwissenschaften or into Latin as scientia. However, a search for scientia in fourteenth-century texts would lead to texts about knowledge. The medieval term for science was natural philosophy (philosophia naturalis). Hence, unless we build into our search engines historical and cultural dimensions, we may find the latest news from the standpoint of one culture but miss finding the richness and complexity of knowledge as a whole. Here again the digital reference room offers a solution precisely because reference rooms include earlier classification systems and multiple definitions of terms.
The lecture will include preliminary examples of a new System for Universal Multimedia Access (SUMMA) which is being developed at MMI. It will also review a series of other tools (including Infobyte's Virtual Exhibitor) and virtual reality reconstructions (e.g., of the Vatican). As a keynote this lecture will provide technologically minded members of the Internet Society with a vision of the enormous challenges and possibilities posed by cultural content.
In the U.S., the rhetoric about the information superhighway was initially focused on pipelines: the engineering infrastructure needed for connectivity. In Europe, where there is a goal of an information society, global interoperability of networks has been recognized as a fundamental precondition. This challenge is being addressed at a European level through various R&D and deployment initiatives and at the world level through G8 pilot projects. These projects are providing the pipelines for high-speed transfer of information. As in the film Field of Dreams, the general approach was "build it and they [in this case, content] will come"; the underlying assumption was that once connectivity is established, the problems are solved, and all one needs to do is to send content down the pipeline.
In addition, we need an intellectual framework for interoperability of contents. Important contributions in this direction are being made by the Internet Engineering Task Force of the Internet Society and the World Wide Web Consortium (W3C), particularly through their Resource Description Format (RDF). Among the North American initiatives also representing useful steps in this direction are those of the National Science Foundation (NSF) in the domains of digital libraries and education, the Coalition of Networked Information, the National Initiative for Networked Cultural Heritage (NINCH), and the Dublin Core in the realm of metadata and ontologies (the American term for thesauri). The first part of the paper reviews major initiatives to provide networked access to cultural heritage with respect to libraries, museums, and education. These initiatives are introducing some common standards to permit interoperability among distributed collections. A fundamental shortcoming of all these solutions, however, is that they are focused almost exclusively on contemporary knowledge and therefore ignore the historical and cultural dimensions of knowledge organization.
The second part of the paper suggests that centralized, digital reference rooms offer a new answer to problems of distributed, networked knowledge. Reference rooms are the traditional equivalents of search engines and structuring tools. Indeed, the reference rooms of libraries have served as civilization's cumulative memory concerning search and structure methods through classification systems, dictionaries, encyclopedias, book catalogs, citation indexes, abstracts, and reviews. Digital reference rooms thus offer keys to more comprehensive tools.
The concept of digital reference rooms is part of a larger vision to create an intellectual framework for interoperability of content. This vision is being developed within the MEDICI (Multimedia for Education and Employment Through Integrated Cultural Initiatives) framework of the European Commission through a new European Network of Centres of Excellence in Digital Cultural Heritage and Information Communication Technology based at Maastricht.
In the early phases, this process was referred to as library automation or electronic libraries. Now the term digital libraries is most frequently used [1]. At the international level, the ISO has a technical committee on information and documentation, including the presentation, identification, and description of documents (ISO/TC46/SC9) [2] and a standard for bibliographic references to electronic documents (ISO 690-2). There is also the G8 Pilot Project 4: Bibliotheca Universalis, which is coordinating the author lists of five national libraries (Belgium, Britain, France, Portugal, and Spain). Japan [3] has its own approach to this G8 pilot project (see Fig. 1).
| Data Retrieval for Source Data | Intelligent Retrieval | Optional Functions | High Definition |
|---|---|---|---|
|
|
|
|
Figure 1. Schema of key elements of Electronic Library System (Ariadne) in Japan's model for the G7 pilot project on libraries.
In addition, the G8 Pilot Project 6, Environmental Natural Resources Management, also has a Digital Library Reference System. [4] The United Nations has a bibliographic information system (UNBIS). [5] The United Nations Educational, Scientific, and Cultural Organization (UNESCO) has a Memory of the World Program [6] and a World Heritage List, [7] which also pertain to museums. A number of international institutions are exploring problems of digital libraries; these include the International Federation for Information and Documentation [8]; International Federation of Library Associations; International Research Library Association; International Association of Digital Libraries, which has a millenium project (re: the New World); and the International Council on Archives, which has an ad hoc committee on descriptive standards.
At the European level, there is a Gateway to European National Libraries. [9] The European Commission's Telematics for Libraries [10] has projects on a number of themes, including children's pages, distance learning, journals, metadata, music libraries, and software. Projects include Automatic Information Filtering and Profiling [11]; Catalogue with Multilingual Natural Language Access/Linguistic Server [12]; a European SR-Z39.50 Gateway [13]; Heritage and Culture through Libraries in Europe [14]; Integrated Library Information Education and Retrieval System [15]; Online Public Access Catalogue for Europe-II [16]; Large Scale Demonstrators for Global, Open Distributed Library Services [17]; and Virtual Library. [18] The Telematics for Research program is also sponsoring the Development of a European Service for Information on Research and Education, [19] which in turn has links to a British initiative on resource organization and discovery in subject-based services (ROADS). As part of the ESPRIT long-term projects, the European Research Consortium for Informatics and Mathematics Digital Library [20] has developed a semantic index system and thesaurus management system (SIS-TMS) and a platform-independent and interplatform multimedia (REFEREED).
In addition, both ESPRIT and INFO2000 projects involve examples of digital library content that include a dictionary of art, palaces, and gardens of Europe; great composers; a multilingual multimedia encyclopedia of ecology; the World Electronic Book of Surgery; and an information context for biodiversity. [21] There is the European Digital Library Consortium. [22] Another consortium, called Multimedia Electronic Memories at Hand (MEMORIA), [23] includes the Oxford Text Archive, the Bibliothèque Nationale, and the Consorzio Pisa Ricerche and is devoted to accessing, retrieving, and structuring writing.
The European Commission's Guide to Open Systems Specifications [24] has sections on information structure and representation as well as library applications. The W3C [25] is concerned with digital libraries. The Research Libraries Group has their Archives and Manuscripts Taskforce on Standards. There is the International Collaboration on Internet Subject Gateways mailing list. [26] Based at the University of Oxford and the University of Illinois (Chicago) is the important Text Encoding Initiative. There is also the International Institute for Electronic Library Research at De Montfort University. [27]
All over the world are many national-level initiatives. In Australia, for instance, the National Library at Canberra has the Australian Cooperative Digitisation Project 1840-1845. [28] In Canada, the Bureau of Canadian Archivists has their Planning Committee on Descriptive Standards and Rules for Archival Description. The National Library in Ottawa has the Canadian Initiative on Digital Libraries [29]; working groups on advocacy and promotional issues; creation and production issues and organizational and access issues (metadata); and digital projects, [30] including Early Canadiana Online, [31] the Virtual Canadian Union Catalogue, [32] and the Virtual Visit. [33]
Denmark has a national project for Denmark's electronic research library (DEF). In France, the Bibliothèque Nationale de la France, affectionately known as the "tgb" (très grande bibliothèque), is engaged in the MEMORIA consortium (mentioned above) and the Gallica [34] project. In Germany, there is work on a distributed digital research library (Verteilte Digitale Forschungsbibliothek). A federation of five libraries (Berlin, Frankfurt, Gottingen, Munich, and Wolfenbüttel) is developing a pilot project for the digitizing of all German works (Arbeitsgemeinschaft Sammlung Deutscher Drücke). This project includes a Kompetenz Zentrum Digital Library sponsored by the German Research Society (Deutsche Forschungs Gemeinschaft). Bertelsmann, [35] one of the largest publishers in the world, is active on a number of fronts.
The Gesellschaft für Mathematik und Datenverarbeitung (GMD) has their Integrated Publication and Information Systems Institute (IPSI), [36] which is concerned with distributed processing of multimedia objects. This group is working on a global electronic and multimedia information systems for natural science and engineering [37] (GLOBAL-INFO), which includes physics (PhysDoc), computer science (McDoc); mathematics (MathNet); and natural sciences and technology (eprint). Associated with these projects are the bibliographies on database systems and logic programming (dblp; Trier) and on computer science (Karlsruhe) and the project on advanced retrieval support for digital libraries [38] (DELITE). The Institut für Terminologie und angewandte Wissensforschung in Berlin is exploring full text digitization and retrieval using Standardized Generalized Markup Language (SGML). The project From Text to Hypertext includes at least six partners: Börsenverein des Deutschen Buchhandels, Schlütersche Verlagsanstalt, Olms Verlag, Bertelsmann Club GmbH, Deutscher Bibliotheksverband e.V., and the Zentrum für Kunst und Medientechnologie. The Library at Braunschwieg is mapping headings in eight major catalog systems. [39]
Italy has a project for an Intelligent Digital Library in Bari and a Biblioteca Italiana Telematica [40] (CIBIT) at the National Research Council site in Pisa. In Japan, the Digital Library Network is based at the University of Library and Information Science (Tsukuba Science City). [41] In the Netherlands, the Digitale Encyclopedie Nederland project foresees the creation of a digital encyclopedia of Netherlandish cultural heritage, which will cover both library and museum materials. The publisher Elsevier, engaged in the German GLOBAL-INFO project (mentioned above) is also involved in the University Licensing Program. [42] The Netherlands is also developing a Dutch Electronic Subject Service. [43]
In the U.K., a number of digital library initiatives are based at the Bath Information and Data Services [44] and the Electronic Library Programme, [45] which entails a range of problems, including access to network resources, digitization, electronic document delivery; electronic journals; online publishing, and quality assurance. The British Library has a digital library research program, [46] a digital library program, [47] and the project Cataloguing and Retrieval of Information Over Networks Applications. There is also the United Kingdom Pilot Site Licence Initiative. The Scottish Cultural Resource Access Network [48] includes both library and museum resources.
The U.S. is a nominal member of the G8 Pilot Project 4: Biblioteca Universalis. [49] At home in the U.S., the military is closely connected with digital library initiatives. The Defense Advanced Research Projects Agency (DARPA) has a research program on national-scale information enterprises and has developed both the Knowledge Query Manipulation Language [50] and the Knowledge Interface Format. [51]
| 1. Carnegie Mellon University | Infomedia: Digital Video Library [52] |
| 2. Stanford University | Integrative Mechanisms in Heterogeneous Services [53] |
| 3. University of California, Berkeley | Environmental Planning and GIS [54] |
| 4. University of California, Santa Barbara | Alexandria Project Spatially Referenced Map Information [55] |
| 5. University of Michigan | Intelligent Agents and Information Location [56] Advanced User Interface [57] Glossary: Terms, Organizations [58] |
| 6. University of Illinois | Federating Repositories of Scientific Literature [59] Social Science Team [60] Semantic Research [61] Interspace Prototype [62] Includes CS Quest Automatic Indexing Concept Space Generation Visualization Fisheye View Systems Software Research Group |
Figure 2. Key aspects of the Stanford Digital Libraries Project, [63] now called the Digital Libraries Initiative. [64]
Along with the NSF and the National Aeronautics and Space Administration (NASA), [65] DARPA is active in the Digital Libraries Initiative [66] (see Fig. 2). DARPA funds the new journal D-Lib, and along with the NSF, it recently (1998) announced the International Digital Libraries Collaborative Research, [67] which is linked with the British Joint Information Systems Centre. The National Information Infrastructure (NII) is developing an NII Virtual Library. [68] The Information Infrastructure Task Force includes a linguistic data consortium (e.g., BBN, SRI, MIT, CMU) working on a spoken natural language interface to libraries.
Indirectly, the U.S. military is also active in four projects of the Department of Energy in the digital libraries domain: the Comprehensive Epidemiologic Data Resource; the Socio-Economic Environmental Demographic Information System; the Carbon Dioxide Information Analysis Center, and a database of scientific mathematical software. It is also indirectly active in the federal government's initiative on computer and information science and engineering, which has projects on information and intelligent systems; advanced networking infrastructure and research; and experimental and integrative activities, which represent eight different areas:
The U.S. government's research on digital libraries includes a library server for manufacturing applications at the General Electric R&D Center, current economic statistics, and the Guide to Available Mathematical Software. Also active are the American Society for Information Science [69] and the Center for Networked Information Discovery and Retrieval. [70] There is the Digital Library Integrated Task Environment, [71] and the Networked Computer Science Technological Reports Library. [72] The Committee on Institutional Cooperation has the Center for Library Initiatives and the Virtual Electronic Library. [73] The Digital Preservation Consortium [74] has the Digital Library Federation [75] and the Machine-Assisted Realization of the Virtual Electronic Library.
The Library of Congress [76] is the seat of the National Digital Library Program, [77] is coordinating the Z39.50 sites [78] (ISO 23950), is engaged in the American Memory Project, [79] and is 1 of 10 libraries in the Ameritech Digital Library Project, [80] which is developing the concept of a digital librarian. The University of Illinois, which is in the Digital Libraries Initiative (see Fig. 2), has four related projects:
The University of California at Berkeley has a series of projects relating to digital library research and development, [85] including an advanced papyrological information system [86]; American Heritage Project; California Heritage Project [87]; Cheshire II Search Service (which uses the Z39.50 protocol) [88]; digital page imaging and SGML [89]; electronic binding document type description (DTD) (ebind); encoded archive description [90] (EAD); Finding Aids for Archival Collections [91]; Index Morganagus [92]; a full-text index of library-related electronic journals; Scholarship from California on the Net [93]; the Berkeley Multimedia Research Center [94]; and the Information People Project. [95]
At least four universities (Berkeley, Duke, Stanford, and Virginia) are engaged in the American Heritage Project, [96] which includes the American Heritage Virtual Archive Project. Cornell University has the Consortium for University Printing and Information Distribution, a project on flexible and extensible digital object and repository architecture [97] (FEDORA), and two projects with the University of Michigan -- the Making of America [98] and the Internet Public Library. [99] The University of San Diego and the University of Southern California are working on the Alexandria Digital Library. [100]
In addition, there are a number of individual initiatives. Harvard University has the Harvard Information Infrastructure Project. [101] Texas A&M University has the Center for the Study of Digital Libraries. [102] The University of Maryland at College Park has the Digital Library Research Group (DLRG). The University of North Carolina at Chapel Hill has been working on the Sharium. [103]
Major companies active in digital libraries include Bell Communications Research (Bellcore), [104] IBM [105] (whose work is focused on four areas: media and entertainment, higher education, [106] government, and cultural institutions); Xerox [107] (whose approach covers five domains: infosphere, workspace, sensemaking tools, document, and superbook); and Lockheed Martin, which is developing a rapid-access electronic library system (RAELS). There are virtual libraries in the biosciences [108], chemistry [109] and control engineering. [110] There are also a number of new journals, including D-Lib, [111] Digital Library News [112] (DLN), Initiatives in Digital Information (University of Michigan), [113] International Journal on Digital Libraries (Rutgers University) [114], Journal of Electronic Publishing (University of Michigan), [115] and the Research Library Group's DigiNews, [116] as well as publishers such as High Wire Press [117] (Stanford University). [118]
At the international level, digital museums, also called virtual museums, [119] or imaginary museums (a term translated from the French musée imaginaire as described by the late minister of culture Andre Malraux), are covered in one of the 11 pilot projects of the G8, namely Pilot Project 5: Multimedia Access to World Cultural Heritage. [120] UNESCO has several projects, including a World Heritage Information Network, [121] World Heritage Web, [122] and Heritage Network, [123] as well as more specialized projects under the Communication, Information, and Informatics Sector, [124] such as Bibliotheca Alexandrina; International Informatics Programme; International Program for the Development of Communication; Memory of the World Program; World Information Report, and World Heritage.
Also interested in these problems are other international institutions, such as the International Council of Museums, [125] with its numerous committees [126] on conservation, documentation, [127] education, monuments, sites and so forth; and the Comité Internationale d'Histoire de l'Art, which set out to create the Thesaurus Artis Universalis (TAU). The Consortium for Computer Interchange of Museum Information [128] has been working on standards and has developed a test bed for museums to share information. The Computer Heritage Information On-Line [129] has developed an exhibition catalog document type description [130] and is working on a standards framework that includes Standard Generalized Markup Language for Cultural Heritage Information [131] and Full-Text Document Type Description V4.0. [132] The Council of Europe has the Division for Cultural Heritage. [133] A group based in Berlin is discussing a new Museum of World Cultures, which would probably be located in Strasbourg.
The European Commission has funded many projects, including several specifically intended to establish networks in the field of cultural heritage. These include Remote Access to Museum Archives; Multimedia European Network for High-Quality Images Registration, which is linked with the commercial enterprise Museums On-Line; Network of Art Research Computer Image Systems in Europe; Sharing Cultural Heritage through Multimedia Telematics (AQUARELLE) [134] and MIDAS Net. [135] There is also an Advanced Communications and Technologies Services project called Virtual Museum International (VISEUM). Museums over States in Virtual Culture [136] (MOSAIC) is part of the Trans-European Networks Project (TEN). In the context of INFO2000, there have been Art Web, the Mediterranean Multi-Media Support Centre for Culture and Arts (M.Cube), [137] the Cultural Heritage of Long-standing Legacy in Open Network (CHAMPOLLION), and the Cultural Heritage and Arts Information Network (CHAIN). The European Commission's Telematics for Libraries has developed a Visual Arts Network for the Exchange of Cultural Knowledge (VAN EYCK). [138] There are also several networks involving pay services, namely, Museums On-Line [139]; Artois, of the Réunion des Musées Nationaux (RMN) [140]; and Art Web; and in the U.S., the Art Museum Image Consortium [141] and Bill Gates' Corbis.
At the national level, the Canadian Heritage Information Network [142] was the first network to be established for museums; it was followed by Germany's Marburger Archiv, [143] from which the Marburger Informations-Dokumentations und Administrations- System (MIDAS) was developed. In recent years, there have been a number of such national networks, including the Scottish Cultural Resources Access Network [144] and the Australian Cultural Network. [145] Such countries as France have created a general inventory of monuments and artistic treasures, [146] a database of Fine Arts and Decorative Arts (JOCONDE), and a database concerning conservation and restoration (MUSES). The RMN [147] has developed a new network of online museum shops. Italy is developing Cultural Heritage Assisted Analysis Tools. The U.K. has an Arts and Humanities Data Service, [148] Museum Documentation Association [149] (MDA [150]), and a National Council on Archives with an Information Technology Standards Working Group and a Royal Commission on the Historical Monuments of England. [151] There is a proposal for a Eurogallery, which will link the National Gallery (London), the Réunion des musées nationaux (Paris), and the Van Gogh Museum (Amsterdam). Also under discussion is a possible European Museum Information Institute.
In the U.S., the American Association of Museums has established a Museum Digital Licensing Consortium [152] (MDLC Inc). National efforts are being led by the President's Committee for the Arts and the Humanities, National Endowment for the Humanities, [153] NINCH, [154] and the Coalition for Networked Information. [155] The American Association for State and Local History has been working on a Common Agenda for History Museums. The Getty Trust, through its (now abandoned) Getty Information Institute, [156] has produced such useful tools as the Art and Architecture Thesaurus, a Thesaurus of Geographic Names, and a Union List of Author Names. The Society of American Archivists has a Standards Board [157] with a Working Group on Standards for Archival Description, a Committee on Archival Information Exchange, and an Encoded Archival Description Working Group. [158] Other significant organizations include the Museum Computer Network, [159] Archives and Museum Informatics (Pittsburgh), and the Association of Art Museum Directors. The University of California at Berkeley also has a Museum Informatics Project. [160]
Major companies in the realm of digital virtual museums include Hitachi (Viewseum [161]), Intel (Virtual Gallery [162]), Sony (Paris), which has a Personal Experience Repository Project [163]; Xerox (Grenoble), which is working on the Campiello Project -- a part of the European Commission's i3 or Icube projects. [164] The University of Karlsruhe is developing Ontobroker. [165]
Similar developments are evident in the fields of education and training. At the global level, the International Telecommunications Union is creating a virtual training center for distance learning (ITU/BDT). [166] There are the Global Telecommunication University, the Global Telecommunication Training Institute, a virtual training center, [167] a global campus [168] (not to be confused with the IBM Global Campus), [169] the Global Learning Consortium, [170] the International Society for Technology in Education, [171] the Federation for Audio-Visual Multimedia in Education, and the Council of European Informatics Societies (CEPIS), [172] which is producing the European Computer Driving Licence. [173] There is also The Association for Computer-Based Training (TACT). [174]
At the European level, there is a database on higher education in Europe, called Ortelius [175]; the Educational Multimedia Task Force initiative [176]; the European Schoolnet [177]; and the European Education Partnership, which is linked with the MEDICI framework. [178] As in the area of culture, the European Commission has produced a Memorandum of Understanding for partnership and standardization in the field of learning technology (MOU Learning). [179] In the U.S., the EDUCOM consortium has introduced EDUCAUSE, [180] and in conjunction with the National Learning Infrastructure Initiative, [181] it is working on an instructional management system [182] "to enable an open architecture for online learning." In turn, IMS is cooperating with IEEE's Learning Technology Standards Committee [183] -- specifically, their Technical Standards for Learning Technology Working Group (P1484.1) -- to create an architectural and reference model of a learning environment. IMS is working with the European Union's Annotatable Retrieval of Information and Database Navigation Environment (ARIADNE) project for the development of information content metadata. [184] IMS is also part of the Office of Science and Technology Policy, which is linked with the Department of Defense and the White House [185] and is developing advanced distributed learning. IMS is also working with the Aviation Industry CBT Committee, [186] which is developing recommendations and guidelines for a common learning environment model.
To address the problem of how these myriad initiatives can be integrated into a larger framework, the European Parliament has recently outlined the First European Community Framework Programme in Support of Culture. [188] This program will integrate existing projects such as Kaleidoscope, [189] Ariane, [190] Raphael, [191] City of Culture, [192] and Media II: Audiovisual Policy, [193] so that they are all included within a single funding structure. [194] The Fifth Framework Programme of the European Commission will reflect these goals. [195]
In 1996, as a part of this strategy toward integration, the European Commission introduced a Memorandum of Understanding (MOU) for Multimedia Access to European Cultural Heritage. This action led to the MEDICI framework in October 1998. [196] MEDICI has four main action lines: intermuseum thematic virtual multimedia exhibitions, education, tourism, and best practice lists. The MEDICI secretariat in Milan is developing a Web site with emerging hardware and software standards, templates for contracts, lists of possible sources for funding, as well as notable examples of products. The MEDICI site serves as an electronic version of a best practice handbook for multimedia in the cultural domain and provides cultural institutions, particularly smaller museums, with reliable information about the state of the art, that is, with working solutions that they could not afford to undertake on their own.
An integral part of the MEDICI framework is a European Network of Centres of Excellence in Digital Cultural Heritage and Information Communication Technology, based at the new Maastricht McLuhan Institute (MMI). This network will be linked with the Scuola Normale Superiore in Pisa, the University of Bologna, and other centers throughout Europe. A basic goal of the network is to create an intellectual infrastructure for interoperability of content. A crucial step in this direction will be provided through the development of digital reference rooms.
As noted above, immense repositories of knowledge are found worldwide, in libraries, museums, archives, and other collections. In the past, access to these materials was limited to visiting an individual site, then visiting another, and so forth. A spectrum of solutions evolved to meet this challenge. At one extreme, there was a quest to collect everything in a single location, which inspired the Alexandrian Library and inspired more recent efforts such as the British Library (London) and the Bibliothèque Nationale de France (Paris). As the owners of these collections discovered, no single building or complex can contain all the books and materials that exist.
At the opposite extreme, a vision of the Internet as a completely distributed system has recently developed. This theory is that once the collections are digitized, a new type of universal access to these materials is possible. In theory, anyone can have access to these materials at any time and anywhere, so that the access to knowledge would be greatly democratized. However rhetorically attractive the concept, it is doomed to failure because there is no way of moving efficiently across myriad and diverse naming procedures without standardized names for persons, things, places, and times. Digital reference rooms are an intermediate solution between these two extremes: centralized reference materials furnish authority lists to provide standardized access to distributed knowledge.
Traditional reference rooms contain a series of tools to help readers find books, manuscripts, and other materials. These materials include lists of terms (classifications, thesauri), definitions (dictionaries), explanations (encyclopedias), titles (library catalogs, book catalogs, publishers' catalogs, and bibliographies), and partial contents (abstracts and reviews). These combined tools serve as the collective search methods of civilization.
In such physical reference rooms, a single problem may take us to dozens, or even hundreds, of different sources as we search through multiple definitions of a term, check how the term is handled by different encyclopedias, and then check the locations of books related to that term. Such locations may be different libraries, national book catalogs, publishers' catalogs, and other specialized collections. Digital reference rooms have an enormous advantage over such traditional reference rooms because we can consult many sources through a single interface.
Some aspects of digital reference rooms are already well known through the many existing digital libraries, which are typically limited to electronic versions of library catalogs and offer lists of authors, titles, key words, and/or subjects, and in some cases, a standard classification system (see Fig. 3). For instance, the Online Computer Library Center is providing access to information through the Dewey for Windows program. It is also exploring the possibility of mapping between various systems.
| Form | Content |
| Authors Titles |
Subjects or Keywords Classifications and Thesauri |
| (Full) Texts, Corpora | |
Figure 3. Four basic access points to full texts according to traditional library approaches reflected in digital library projects.
Initial steps in the direction of digital reference rooms can now be found on the Internet. Search engines such as Yahoo and Excite, [197] and sites such as Webdata [198] and Link2go, [199] have reference sections, which include titles and sometimes contents of dictionaries, directories, encyclopedias, thesauri, and bibliographies, as found in reference rooms. The World Wide Web Virtual Library includes such materials under the heading Information Management. [200] The National Center for Supercomputing Applications has a meta-index of Internet resources. [201]
The University of California at Berkeley is working on "webliographies." [202] The University of Strathclyde has created BUBL (originally an acronym for Bulletin Board for Libraries), which has a significant reference section and allows one to search using Dewey Decimal Classification (DDC). [203] The University of California at Irvine has a Virtual Reference Collection. [204] The University of Sussex has a useful list of subject resources for the arts, social sciences, and sciences. [205] Gerry McKiernan at Iowa State has a list of online classification systems and controlled vocabularies. [206] Professor Beard at Bucknell University has made a list of approximately 800 dictionaries [207] and 150 grammars. [208] The Bayerische Staatsbibliothek is scanning in standard German encyclopedias such as Adelung and Zedler. General reference tools such as the International Bibliography of Periodical Literature (Internationale Bibliographie der Zeitschriftenliteratur aus allen Gebieten des Wissens or IBZ) are already available on CD-ROM. The Oxford English Dictionary is scheduled to be available online in October 1999. [209]
The challenge is to combine such sporadic examples and create a digital reference room. This electronic reference room will begin as a personal digital desktop with a handful of standard reference works in electronic form and will gradually expand its repertoire of sources to accommodate the local problems of students and experts and then so the same at national and international levels. Eventually, the reference room will provide an electronic equivalent of the combined reference materials available in the world's great libraries, such as the Bibliothèque Nationale, the British Library, the Library of Congress, the Herzog August Bibliothek, and the Vatican. With the use of such methods as the resource description format of the W3C, the quality and level of acceptance of these sources can also be built into the system.
If digital reference rooms provided only some of the contents of physical reference rooms in digital form, they would be of limited interest. Potentially digital reference rooms offer much more: they provide further resources for authority lists; new search strategies via related terms; new forms of metadata; and keys to historical and cultural dimensions of knowledge. Thus, they are important as a basis for an intellectual framework for interoperability of content. Digital reference rooms are therefore much more than a key to digital libraries and museums. They offer a new approach to the organization of knowledge.
In searching for authors or titles, library catalogs offer an obvious point of departure. However, such catalogs are limited to the authors and titles that happen to be in that particular library. Networks of library catalogs such as the Online Public Access Catalogue for Europe-II [210] or that of the Research Libraries Information Network have greatly expanded such lists of authors and titles.
Reference rooms contain many other resources that can help in the creation of comprehensive lists of authors and titles. These resources include biographical dictionaries, encyclopedias, book catalogs, and bibliographies. Authority lists provide us with standard spellings of names. Reference works such as the Allgemeine Künstler Lexikon provide us with numerous variants of those standard spellings. When we move to full-text searches, these variants become new elements for searching. Thus, digital reference rooms greatly increase the potential number of authors and titles to be searched.
As noted above, museums and galleries are already creating digital versions of their collections, including their reference works. We can therefore use existing digital versions of reference works (e.g., classification systems, dictionaries, encyclopedias, catalogs, bibliographies, reviews, and abstracts) to create new authority lists of persons (Who?), things (What?), places (Where?), times (When?), processes (How?), and causes or reasons (Why?). Some of these lists have been produced by publishers of standard reference works (e.g., Saur). Publishers will find it useful to cooperate in this venture to provide new links to their materials, which can be licensed within the system.
As noted in Figure 1, digital libraries are typically limited to two types of questions: Who? (authors) and What? (titles, subjects, and classifications); thus, they are limited to four points of access. Some libraries, such as the Herzog August Bibliothek (Wolfenbüttel), also offer access via two further questions: Where? (places of publication) and When? (dates of publication).
Digital reference rooms potentially offer a much richer set of entry points to information. The authority lists of authors' names can be linked with corresponding names in biographical dictionaries, encyclopedias, bibliographies, partial contents, and full contents. Similarly, the subject terms can be linked with classification systems, and in turn with dictionaries, encyclopedias, and the like. This linking will have two fundamental consequences.
| WHO | WHAT | WHERE | WHEN | ||
|---|---|---|---|---|---|
| 1. Names, Terms | Classifications | Authors | Subjects | Places | Periods |
| 2. Definitions | Dictionaries | " | " | " | " |
| 3. Explanations | Explanations | " | " | " | " |
| 4. Titles | Titles | " | " | " | " |
| 5. Partial Contents | Abstracts | " | " | " | " |
| 6. Full Contents | (Full) Text or Corpora | ||||
Figure 4. Different levels of knowledge in the digital reference room (1-5) coupled with four basic questions to produce a matrix of 20 different access points into existing corpora.
First, the links will help to contextualize knowledge concerning a person or subject. If I encounter a name and am uncertain whether this is an individual who interests me, I need only to check a short description in a biographical dictionary, or a longer entry in an encyclopedia, to ensure that this is the person I am seeking.
Second, this process of contextualization will simultaneously provide me with a wealth of new material that could be crucial to my searching. For example, when I look for a given word within a classification system, I am given all of its related terms, which provides me with a vocabulary for further searches. Similarly, when I move from the name of a given author to a biographical dictionary, I find the names of key individuals associated with the one who interests me. These names can, in turn, provide me with further information about the individual in question. Taken to its logical conclusions, this approach provides me with at least 20 points of entry into the contents found in full texts or corpora (see Fig. 4).
Standard classification systems, such as the DDC system, allow limited moves between broader and narrower terms. Technically, this is a form of subsumption, which constitutes one of several relationships among terms. The past generation has seen a dramatic rise in new types of thesauri, which differ from earlier classification systems in that they establish a number of relationships among concepts. In a seminal article, Perrault introduced a method of integrating these relationships within the Universal Decimal Classification (UDC) system. The Medical Subject Headings system has five types of relationships. Systems such as DDC are too primitive to allow a full range of relationships. Nonetheless, if there were mappings to link LC and DDC to UDC, then one could use the relationships of UDC as a starting point for links to other systems. These same reference works can be used to generate an enormous amount of related terms in the form of alternatives, associations, complementaries, duals, identicals (synonyms), and opposites (antonyms).
Categories under which titles of books are listed are not limited to classification systems and thesauri. National book catalogs (such as Books in Print in the U.S. and the U.K., Kayser in Germany, Lorenz in France, and Pagliaini in Italy), publishers' catalogs, and bibliographies provide a wealth of further categories. These categories also change over time. Hence, we need to take a given title and trace the various headings under which it is listed to arrive at the equivalent of an etymology of categories. By combining this process with the basic titles in historical bibliographies, we can develop new ways of tracing both the evolution of a field and the changing array of subjects with which it is connected.
Metadata, one of the new buzzwords of digital library initiatives, typically serve as summary data about data that provide a first clue about the authors or subjects of electronic documents or databases. Digital reference rooms introduce the possibility of creating new forms of metadata.
To learn about an individual such as Leonardo da Vinci, I want to know not only the standard and variant spellings of his name but also to be given a standard list of his paintings, drawings, and manuscripts. Hence, a metadata package on Leonardo da Vinci will give me key information on what he produced. At the same time, it will provide me with both regular and variant spellings. Whether I type in "Mona Lisa" or "La Gioconda," the metadata package will direct me to the same painting in the Louvre and will draw my attention to copies and versions available elsewhere.
At present, the Internet is focused primarily on the latest news, as though it were an electronic newspaper. The creators of search engines assume that isolated keywords, presented in the form of natural language and translated into different languages, are all that is required to solve users' problems. Natural language does not account for variant spellings. More important, it does not account for historical and cultural shifts in the meanings of terms.
We cannot reach a deeper understanding of an individual or a subject until we recognize to what extent knowledge itself is a historical phenomenon. A Leonardo expert such as Pedretti now lists more than 40 writings by Leonardo. In 1600, 1800, or even 1900, the numbers of known manuscripts by Leonardo were considerably different. The same holds true of his paintings. Hence, metadata about Leonardo will eventually include more than the titles found in a current catalogues raisonnée or in standard works. It will include lists that change in size with time.
In the particularly complex cases of authors such as Aristotle, the metadata will undergo enormous changes over time. The corpus of Aristotle's works as provided in the Ross edition is very different from the works known at Charlemagne's court in 800; in Paris at the time of Thomas Aquinas' death in 1275, or in Marburg in 1527, when Melanchthon founded the university there.
In the case of countries, the problem of authority lists is more complex still, and present-day search engines ignore the problem entirely. If, for instance, I search for "Poland" in Yahoo or AltaVista, it is assumed I mean contemporary Poland. Historically, however, not only does the name of a country change with time, but its boundaries also change. If I ask about Poland in the fourteenth century, I am asking about a much greater area of land than in 1998. Such information can be found in historical atlases in reference rooms and the in reference sections of map rooms. This information needs to be added to the metadata lists of our knowledge concerning place names.
Not only do basic names, terms, and places change with time, but so do the concepts and structures with which we organize and class our knowledge. It is no secret that science today is very different than it was 600 years ago. The English term science is easily translated into the Latin scientia. However, if we use the term scientia to search among fourteenth century sources, we find references to knowledge. In the fourteenth century, science was classed under philosophia naturalis. The efforts of the Online Computer Library Center to use the DDC system for searching the latest materials in libraries and on the Internet are important for finding subjects according to today's categories but do not help in historical searches.
Throughout the Internet community and even in the museum and library world, there is a general assumption that if only we had "proper" natural language interfaces, combined with faster computers that could potentially do comprehensive full-text searches, all our searching problems would be eliminated. This scenario may be tempting, but it is fundamentally wrong. Simply applying the principle of number crunching to word crunching is not a solution. Unless we understand the historical dimension of knowledge, we shall not even know where to begin with our searches. We may find the latest news from the standpoint of one culture but may miss finding the richness and complexity of knowledge as a whole. Here again, the digital reference room offers a solution precisely because reference rooms include earlier classification systems and multiple definitions of terms.
On the Internet are such recent developments as the Hotsauce software of Apple, which shows a basic term surrounded by all the links at the next level, an approach that has been rendered in three-dimensional form by Stuart Card's Cone Tree and Mathhias Hemmje's Lyberworld. In the future, we need a new type of etymology of terms that will show the changing constellations of its related concepts over time. This area represents another fundamental dimension to which digital reference rooms can contribute. They can help to trace the history of terms (etymologies) and the changing structures of enduring knowledge through catalog headings, classification systems, thesauri, and other reference materials. Thus, digital reference rooms will add a historical dimension to Internet techniques.
Closely related to the above discussion are cultural dimensions of knowledge. These dimensions change over space and are often more subtle than interpretations that change over time (historically). Poland may officially change in size over the ages. Nevertheless, even at a given time, the size of Poland according to the Poles may vary considerably from the size of Poland according to the Germans, Ukrainians, or Russians.
This applies not only to definitions of size, but also to perceptions of meaning or importance. The significance of Siena according to the Sienese is quite different from the significance of Siena according to the Florentines. Eventually, we need metadata to help us view people, objects, and events from the perspective of different cultures. Metadata can help us understand how something that seems wonderful to one person can be horrible for another -- for instance, how the color white, which means purity in one culture, can mean death in another. Only through such means will we slowly develop a framework for a new level of international understanding and, it is hoped, tolerance.
As becomes evident from the above discussion, digital reference rooms have at least six interrelated functions:
The above outline of the interrelated functions of the digital reference room represents a first step toward development of an intellectual framework for interoperability of contents. The MMI -- a European center for digital culture, knowledge organization, and learning technologies -- will combine a prototype of such a digital reference room with new interfaces through a System for Universal Multi-Media Access (SUMMA). A number of other elements are needed to create a comprehensive intellectual framework for interoperability of content. These include problems of conservation and restoration or of reconstruction; more detailed study in the realms of multicultural and historical terminology and classification; accepted versus emerging concepts; and dynamic versus static knowledge.
The new Network of Centres of Excellence in Digital Cultural Heritage will focus on creating this intellectual infrastructure. The network as a whole will have six other functions, as follows:
The links with museums and libraries can thus ensure new sources for European culture through better access to its cumulative memory of creative traditions. The MOU for multimedia access to Europe's cultural heritage was developed in parallel with the original aims of G8 Pilot Project 5: Multimedia Access to World Cultural Heritage. The remaining challenge is to ensure that European experiences are properly seen around the world within the context of culture. For this reason, links with international bodies such as UNESCO and International Council of Museums are being established.
Global interoperability of networks has rightly been recognized as a fundamental prerequisite for an information and knowledge society. In addition to such pipelines, the exchange of information and knowledge requires an intellectual infrastructure for interoperability of content. The first part of the paper reviewed major projects at the international and national levels in the realms of digital libraries, museums, and education. This review showed that myriad useful attempts are under way in these fields -- often, alas, in ignorance of efforts elsewhere. It also brought to light a fundamental shortcoming of all of these solutions: they are focused almost exclusively on contemporary knowledge and, as such, ignore the historical and cultural dimensions of knowledge organization.
In response to a profound need for a whole that is greater than the sum of its parts, the European Parliament and the European Commission are developing a more systematic approach. The Fifth Framework Programme reflects this goal in the context of information society technologies as a whole. In the realm of culture, this goal is reflected by the MEDICI framework.
Within the MEDICI framework, the European Commission is developing a new Network of Centres of Excellence in Digital Cultural Heritage, based at the new MMI. Reference rooms are proposed to serve as the historical equivalents of search engines for the collective memory of civilization. Therefore, digital reference rooms offer a key to the challenge of integrating historical and cultural dimensions of knowledge. Their creation will serve as a first step toward developing an intellectual framework for interoperability of content. The network includes European centers. The goal is to expand this network so that MEDICI's efforts for multimedia access to Europe's cultural heritage can be collaborative with G8's larger goal of multimedia access to world cultural heritage.
At the heart of this quest lies an interesting paradox. To create a truly global network of information and knowledge, we need coherent standards: common rules for recording, storing, transporting, and accessing knowledge. However, the result of this homogenization of forms must be a diversification of contents, that is, a fundamental increase in awareness of the uniqueness of each local area. McLuhan's term global village referred to an emerging reality whereby persons all around the world are linked electronically as if they were in a village. The term must not come to mean that we are all reduced to a single undistinguishable and undistinguished mass. The new technologies must increase our awareness of individual uniqueness. That is why historical and cultural dimensions of knowledge provided by digital reference rooms are not a luxury. Digital reference rooms are a key to more than digital libraries and museums. In addition, they may well be our only hope of realizing the deeper goals of the Internet as a whole: creating a world wide web of knowledge that makes us richer as individuals.
