LEAF-VN HARVESTS OUR FIRST SUCCESS: Bo Quy Tac Bien Muc Anh-My Rut Gon, An Ban 1988, will be soon handed to our colleagues in Vietnam.

             “… We spoke of creating a bridge to connect Vietnamese expatriates … to work with library leaders in Vietnam … cooperated and interacted with Vietnamese librarians in a true spirit of team work.”  This statement from my report in Spring 2001 is affirmed by Ms. Tran Kim Thu, Head of Cataloging Department of the National Library of Vietnam (NLV), Hanoi, in her recent email, “Thua chi Lien Huong:  Trong cung mot ngay toi  va chi da  gap nhau tren  cung mot  mang, du  rang ve   mat dia-li, thi  qua xa la voi-voi. Rat cam on chi da la cau noi giua toi va cac anh chi (ca nguoi Viet va nguoi My) trong cung mot nghe chung.” (Dear Lien-Huong:  Within the same day, you and I are truly on the same side although geographically how far away we are.  Thanks so much for serving as a bridge to connect colleagues (Vietnamese as well as American) of the same profession).

            This teamwork among the members of LEAF-VN and Vietnamese librarians has excelled and fruited during fiscal 2002: the Bo Quy Tac Bien Muc Anh-My Rut Gon, An Ban 1988, (The Concise AACR2, 1988 Revision) will soon be handed to our colleagues in Viet Nam.  As the Representative for our Board of Directors, I am expressing my deep gratitude for those who have made this Project a reality.

              The Project started in late 1999 with the volunteer work late into the nights of the two translators, Mr. Lam Vinh-The and Ms. Pham Thi Le-Huong.  In November 2001, the review process began with the translators and the National Library of Vietnam Review Committee.  The fundraising effort lasted for two years!  Although, LEAF-VN acknowledged the support of donors in the Bo Quy Tac Bien Muc Anh-My Rut Gon, An Ban 1988, on behalf of our Board of Directors, I sincerely thank these generous supporters again!  The following people whose names should be mentioned here for the success of this translation project are: the two translators, Ms. Pham Le-Huong and Mr. Lam Vinh-The; the author of the Concise AACR2, 1988 Revision [Concise Anglo-American Cataloging Rules, 1988 Revision] , who also wrote the Introduction for the translated manual, Mr. Michael Gorman; Ms. Tran Kim Thu, Head of the NLV Review Committee, Mr. Vu Van Son with his numerous editing for the manuscript, the NLV Director Pham The Khang with his enthusiastic support; and Mr. Ngo Duc Diem, President of VAP Consulting, LLC of San Jose, California, whose large contribution completed the fundraising! 

               On November 13^th, when I deposited Mr. Ngo Duc Diem's contribution I realized that LEAFers could truly celebrate!  I sent him a celebrating card beside LEAF-VN’s IRS-Thank-You Letter.  He deserves a bottle of champagne from us!   We reached the finish line after a fundraising marathon.  

Introduction

It has often been observed that the current Web is chaotic for finding information. It needs help and we can provide it! 

Introducing an element of authority control to the Web environment would help meet these objectives:

• Assure consistency
• That in turn would enable us to efficiently collocate works in our catalogs through the use of controlled forms of headings
• It would facilitate the sharing of the workload to reduce cataloguing costs. We hope authority files could be shared among all communities - libraries, archives, rights management agencies, publishers, and others.
• Shared authority information has the added benefit of reducing the global costs of doing authority work while enabling controlled access and better precision of searching.

Other objectives for authority control are

• to enable users to access information in the language, scripts, and form they prefer or that their local library provides for them and
• to simplify the creation and maintenance of authority records internationally.

The virtues of authority control have been debated and restated for decades.  When we apply authority control in the Web environment, we are reminded how it brings precision to searches, how the syndetic structure of references enables navigation and provides explanations for variations and inconsistencies, how the controlled forms of names and titles and subjects help collocate works in displays, how we can actually link to the authorized forms of names, titles, and subject that are used in various tools, like directories, biographies, abstracting and indexing services, and so on.  We can use the linking capability to include library catalogues in the mix of various tools that are available on the Web.

Controlling forms used for access and displays provides consistency for users.

We are all aware of very poor OPACs that lack cross references or links to authority files and without these features, quite frankly, they are not Catalogues!

Over the past few years there have been several projects that help us get closer to providing authority control on a global scale.  There are several projects sponsored by the European Union, such as the AUTHOR Project that converted a sampling of authority records from the 7 participating countries to the same communication format, UNIMARC.  The LEAF project is looking at linking authority files for archival purposes using Z39.50 protocols and possibly OAI protocols.  The <indecs> and INTERPARTY projects were looking for cooperative work among libraries, museums, archives, and rights management communities in sharing authority information.

Within the International Federation of Library Associations and Institutions, the IFLA MLAR (Minimal Level Authority Records) Working Group identified essential data elements needed in authority records (today we’d call these metadata).  This work continues by the IFLA Working Group on FRANAR (Functional Requirements for Authority Numbers and Records.  They are reviewing and updating the MLAR findings and recently enlisted the help of Tom Delsey in extending the FRBR model (IFLA’s Functional Requirements for Bibliographic Records) to authority records.

Within the digital metadata community, there is a Dublin Core “Agents” working group that continues to explore recommendations for dealing with authority information in the digital environment.

There is also a newly created DELOS/NSF working group on “Actors-Roles” that is exploring controlled access to names of persons, corporate bodies, and automata!

At OCLC discussions continue about CORC authority records - This is an OCLC project that looks towards global expansion to build an authority file.  CORC now provides simultaneous creation of both MARC 21 and Dublin Core bibliographic records records.

Another development over the past few years has been the acceptance of Unicode - it’s now used in the Microsoft tools, under the Windows operating system, which has given it a wider application and visibility.  Unicode itself facilitates more global compatibility through the ability to create and display essentially all scripts for all languages. 

There is also the expansion of NACO and SACO to users of the Anglo-American Cataloguing Rules and Library of Congress Subject Headings worldwide that is promoting authority control on a global scale.

The advances with Unicode, testing of projects on the Web, and new  technological capabilities are all coming together now.  We are really at the brink of making a virtual international authority file a reality!

Universal Bibliographic Control – A New View

We’re also making an historic change to how we view Universal Bibliographic Control (UBC).  The IFLA UBC principles for authority control are parallel to those for bibliographic control, namely that

-         each country is responsible for the authorized headings for its own personal and corporate authors (they didn’t mention uniform titles, series, or subjects), and

-         the authority records created by each national bibliographic agency would be available to all other countries needing authority records for those same authors.  Even more, that the same headings would be used worldwide. 

In the 1960’s and 1970’s when this was really catching on, technology had not yet advanced to make such sharing practical on an international level.  Plus the lack of funding for an international center to manage such a program prevented that visionary concept from becoming reality.  As for the same form being acceptable worldwide, the IFLA developers at that time were primarily from North America and Europe and apparently did not acknowledge the necessity for multiple scripts.

For the past couple of years a new view of Universal Bibliographic Control is emerging from several working groups within IFLA.  This new perspective reinforces the importance of authority control, yet puts the user first…It’s a practical approach that recognizes a user in China may not want to see the heading for Confucius in a Latinized form but in their own script.  Similarly users in Japan or Korea would want to see the heading in their own script and language.

Yet to still get the benefits of shared authority work and creation of bibliographic records that can be re-used worldwide, we can link authorized forms of names, titles, and even subjects through the authority files of national bibliographic agencies and other regional agencies to create a virtual international authority file. 

These are several models for how this might work and we need to do more pilot projects of prototypes of thee models to test which would be best to pursue.

In order to be of most use to the library users in each country, the scripts should be the scripts they can read!  The following figure

 shows that the names we give to an entity can be expressed in many languages and in many scripts.  For example, we could write it in English or German with a roman script, in Russian in Cyrillic scripts, or in Japanese (in several scripts!) and in many other languages and scripts. 

Transliteration may serve as a way for some users to be able to decipher records, but much better is the accuracy of using original scripts. 

We should now provide at least cross references for variant forms of headings in variant scripts when that is appropriate.  In the United States the group within the American Library Association responsible for changes to the MARC 21 format, known as MARBI, is starting to explore this possibility.  More work needs to be done. We should eventually be able to display the script and form of a heading that the user expects and wants.

 I believe that many catalogers within IFLA realize the value of preserving parallel authority records for the same entity.  This allows us to reflect the national and cultural needs of our individual users, and at the same time to allow us to set up the syndetic structure of cross references and authorized forms of headings to be used in our catalogues intended for a specific audience.  It also allows us to include variants in alternate scripts, at least as cross references for now.

Some local systems already provide us with computer-assisted mechanisms for

automatic checking of headings against an existing authority file, and we could see this expanded to then launch a search against a virtual international authority file, if no match was found locally. 

We can also envision the capability of displaying the found matches from the virtual file for a cataloguer to edit or to merge information, if desired, into the local authority record, including capturing the information for future linking.

Some systems now provide community specific retrievals to concentrate on the subject needs of a community in selecting resources for online searches, and other systems like “my library” or “my opac” even go beyond that to individual specific retrievals.  Those could build in the authority preferences for user preferred scripts and displays for controlled vocabularies. 

We want to have the authorized form preferred by a library as the default offered to most users, but we can also envision offering user-selected preferences through client software, or cookies that let the user specify once what their preferred language, script, or cultural preference is - for example for spelling preferences when cultures have variations, like American English and spelling preferences in the United Kingdom - labor and labour...

Let’s quickly take a look at a scenario for how computer systems can help us in the future with authority control on an international level.  A cataloger types in information.  The local system checks the local authority file and finds no match, so it tells the cataloguer that the heading was not found and launches a Web search to the virtual international authority file. 

Up pops the match with a record created at the National Library of Russia in St. Petersburg. 

Our cataloguer takes a look and perhaps doesn’t want all the information but likes a reference or two and wants a link, so

 the local system asks the cataloger if she wants the system to create a basic authority record from the one found and to make a link to it…and the cataloger clicks on “yes.”

The local system then automatically builds a local authority record, grabbing the linking information from the virtual authority file - that is the record from St. Petersburg, Russia.  The cataloger then adds the MARC field 100, authorized form, according to the locally used cataloging rules, in this case AACR2. The cataloger can add other fields if needed. 

The local system adds the linking 700 field - the MARC format has the 700, 710, and 711 fields in authority records, where we can put the linking authorized form and the record control number and the source information for future linking. This linking of authority files would primarily be among the national or regional authority files of national bibliographic agencies - depending on the model we choose.  I’ll come back to that in a minute.

So we’ve now added another link in the virtual international authority file to the authorized form following AACR2  - note the record control number for the Library of Congress: (LC) n79072979 - and the Russian record for the same entity following the Russian cataloguing rules in Cyrillic script - note the record control number from the National Library of Russia: #10326. 

Then our local system updates our local bibliographic record.

When a user comes along, the local system or the “cookies” on the user’s system, could specify they want to see the Cyrillic form and we could display it for them. 

You can also imagine displaying any script, or a Braille keyboard output, or we could provide voice recognition response, built on a user’s profile or their  “cookie.” 

Let me show you how this might look applied to an authority record for Confucius.

This is an example of what a Library of Congress authority record might look like with Unicode capability to include original scripts as cross references in a library’s catalog.  Actually with Unicode the roman script diacritics would appear with the letter rather than before the letter shown here, but this just gives you an idea of what it would be like. The Library of Congress, that follows AACR2 and prepares catalog records in the English languages, could provide cross references in other scripts and languages when that was appropriate. Here we see that the authorized form is Confucius - the well-known English form of the name found in standard English-language reference sources.  There is also transliteration of Chinese, Japanese, Korean, and other language forms for his name, and cross references are also given for those variations in the original scripts - here showing just Chinese, Japanese (katakana and hiragana), and Korean hongul.

There is no particular order to the arrangement of the references, except to place the non-roman scripts following the roman scripts, but that is purely arbitrary.  This model shows English, German, Italian, Chinese, Japanese, Korean, Russian, and transliterations (including Wade-Giles and pinyin for the Chinese, since the Library of Congress just switched to use pinyin).  

This also shows the use of a linking 700 field to show that an authority record was located at the National Library of China and the form of authorized heading according to their rules.  Our system could bring that information in from Beijing after we searched the VIAF and found their record. Or we might get it from the Hong Kong Consortium HKCAN. 

Notice also the new MARC 21 capability to include the URL for a Web page in the last 670 note field.

VIAF Models

To get the benefits of shared authority work and creation of bibliographic records that can be used collaboratively worldwide, we can link authorized forms of names, titles, and even subjects through the authority files of national bibliographic agencies and other regional agencies to create a virtual international authority file.  There are several models for how this might work, and we need to do more pilot projects to test which would be best to pursue.

If we agree that sharing authority information on a global scale is worthwhile, how do we get there? 

Several major authority files exist, built according to their own cataloguing rules and rule interpretations.  We need a one-time project to link the existing records for the same entity - a retrospective matching project.  One suggestion has been to use matching algorithms, such as those developed by Ed O’Neill and others at OCLC, building on bibliographic clues for machine matching at a fairly high level of accuracy.  A “proof of concept” project to test this approach is planned this year between OCLC, the Library of Congress, and the Deutsche Bibliothek (German national Library) in Frankfurt, Germany.  We would still have manual matching and checking to do, but expect machine matching will be a great help. 

We could also have the computer add linking text strings and record control numbers or an entity identification number to facilitate later links and pathways to preferred forms for displays.

This shows a centralized model for a Virtual International Authority File.  We may find that this model is a good approach in terms of record maintenance.  It uses the Open Archives Initiative (OAI) protocol to “harvest”  metadata from existing authorized authority files.  The harvested data resides on one or more servers and is refreshed whenever changes are made in any of the national files.  This means the day to day record maintenance activities continue to be managed as they are now by the National Bibliographic Agency (or regional authority).  We also want to build in the linking, to assure a high degree of precision in the searching in this model; and there are ways to include the links for entities in this model. 

We can also envision a shared international authority file being an integral part of a future “Semantic Web.”  You may have heard about this in the recent Scientific American article by Tim Berners-Lee, founder of the Internet.  The idea is to make the Internet more intelligent for machine navigation rather than human navigation of the Web.  It involves creating an infrastructure of linked resources and the use of controlled vocabularies, they are calling “ontologies.”  These ontologies could be used to enable displays in the user’s own language and script.

Here’s where libraries have an opportunity to contribute to the infrastructure of the future Web - we already have controlled vocabularies in our various authority files.  Those would be linked with other controlled vocabularies of abstracting and indexing services, of biographical dictionaries, of telephone directories, and many other reference tools and resources to help users navigate and to improve the precision of searches, so users could find what they’re looking for.

All of these tools would also link to their respective databases for bibliographic and other resources.  For example, the Library of Congress authority files would link to the bibliographic and holdings databases of the Library of Congress and even to our digital repositories for the linked digital objects themselves. We would also build in the search engines and future tools that, as a collective resource, would connect us to the entire digital world.

This is not going to happen over night, but it is the direction we are headed. 

It’s fun to think about the possibilities and opportunities for testing this out and to think about how we can improve upon our dreams.

The Web has brought us a new way to convey information. The new twist is that our catalog - that is our PC where the online catalog is displayed, is also the device for viewing the actual digital objects and connecting to the entire digital world.   

This gives you ideas of how catalogers can build authority records on the Web and then, once the authority control structure is there worldwide, this can include other stakeholders (publishers, rights management agencies, archives, museums, and other libraries) - all  can use this information and reduce costs.