Articles
spacer

D-Lib Magazine
September 2000

Volume 6 Number 9

ISSN 1082-9873

Directories of Libraries and Related Organizations

 

Judith Pearce
National Library of Australia
[email protected]

Red Line

spacer

Summary

This paper describes data standards work currently in progress to support the deployment of directories of libraries and related organizations in a networked information services environment. Several interconnected directory-related projects are described that are contributing to development of these standards through the implementation of services that address gaps in the Australian resource sharing infrastructure. The relationship between this work and other data modeling initiatives is reviewed and some interesting conclusions drawn.

Introduction

A quick glance under "Directories" in the Digital Librarian [1] reveals that numerous directories of libraries and related organizations are available on the Internet, ranging from lists of web sites and web-accessible catalogues to detailed entries that may have been collected originally for a print publication. Their purpose is to enable the discovery and location of the organizations themselves, as well as information about their collections and services.

Directory information like this has an important role to play in a networked information services environment. However, for this role to be fully realized the information will need to be made accessible through standard protocols that can extract data relevant to a given process. Moreover, the information itself will need to be shareable to minimize the need for the same updates to be reported by an organization to multiple directory services. Currently, a library may need to register with and provide similar details to multiple online directories as well as to applications such as bibliographic utilities, union catalogues, interlibrary resource sharing systems, Conspectus databases and directories of Z39.50 targets.

Definitions

In this paper, the term directory is used to mean a collection of information about parties. The term party is taken from the Indecs data modeling initiative [2]. It is used here to mean a person, or organization that may perform a range of roles over time as an agent in the making and use of resources. The primary parties of interest are libraries and other information service providers, museums, archives, commercial document suppliers, resource sharing consortia and interlending utilities. The primary role of interest is that of resource provider.

The term resource is used to mean any item that is a manifestation of an expression of a work as modeled in IFLA�s functional Requirements for Bibliographic Records (FRBR) [3].

The term Collection is used in the sense defined by Heaney [4] to refer both to aggregations of resources and to aggregations of resource descriptions. The National Library of Australia�s manuscript collection is a collection of resources. The Register of Australian Archives and Manuscripts (RAAM) is a collection of resource descriptions.

The term Service is used in the sense defined by the PRIDE Project [5] to refer to a system (manual or automated) that provides access to a collection. The web gateway to the Register of Australian Archives and Manuscripts [6] is an example of an online service accessible through the HTTP protocol. The RAAM Z39.50 target currently under development will be another online service providing access to the RAAM collection through the Z39.50 protocol. Interlibrary loan / document delivery services providing access to resources described in RAAM are examples of offline services.

Union catalogues as directories

RAAM is often thought of as a directory because it is a union catalogue. A union catalogue describes resources held by more than one organization, by linking resource descriptions to the holding organization through one or more holdings statements. In the Z39.50 Holdings Schema [7] this is done through a SiteLocation entity which is used to report identity and location information about the holding organization or site. In addition, data elements in the schema have been defined that allow collection use and access policy information (including service and copyright fees) to be included as part of the holdings statement where the data is available or appropriate.

At the simplest level of reporting, the holdings statement need only be a link from the resource description to an organization identifier such as a union catalogue symbol. Higher levels of reporting or display may require a client or server to be directory-enabled. By this is meant the ability to query collections of directory information that may or may not be maintained on the same system, in order to obtain information needed to support a given process.

As well as populating holdings statements, there are other use cases needing to be supported in the union catalogue environment that depend on searching directory information. These include the need to limit a search of a union catalogue (or the display of holdings statements or a list of potential resource providers) to holding organizations in a given region or of a given type or belonging to a given consortium or supporting a given service level.

Australian directory projects

In Australia there is a well developed resource sharing infrastructure based on contribution of holdings information to a national bibliographic database and conformance with an Australian interlending code. Key players in the Australian library community had been aware of the need to address the role of directory services in this infrastructure for some time. It emerged as a priority task following completion of several projects that have significantly enhanced the technical infrastructure. These include:

  • Implementation of an Australian Libraries Gateway with an online update capability enabling libraries to take responsibility for maintaining their own entries.
  • Migration of the Australian National Bibliographic Database to a Z39.50 server (AMICUS).
  • Migration of the national interlibrary lending utility (Kinetica Document Delivery) to a system supporting exchange of requests via the ISO ILL protocol (Fretwell-Downing's OLIB VDX).
  • Customisation for deployment within the Australian university sector of a Local Interlending and Document Delivery Administration System (LIDDAS) based on the ISO ILL protocol and Fretwell-Downing's OLIB VDX.

For more on these projects, see Mclean [8].

Work is also proceeding to address the access management aspects of information service delivery in the academic and research environment. The PRIDE Project has been testing a directory-enabled systems architecture that will support the development of local information service management systems providing unmediated access where possible to a range of local and external information services.

Kerry Blinco [9], Macquarie University's representative on the PRIDE Project, has developed an Information Services Model which illustrates the role of directory systems in this architecture [Figure 1].

Figure showing Information Services Model illustrating the rold of directory systems in this architecture.

Macquarie University has successfully used this model as a tool for identifying and addressing gaps in the technical infrastructure in projects such as the JEDDS [10] and LIDDAS projects [11]. To further the goals of the PRIDE Project, they are collaborating with the National Library of Australia to set up an X.500 test bed server containing directory entries for the libraries participating in the Australian national resource sharing network. The information in this directory will be derived from entries maintained online by the organizations themselves through the Australian Libraries Gateway [12].

The Australian Libraries Gateway is being extended as part of this process to incorporate data currently stored in a separate Interlibrary Resource Sharing (ILRS) Directory [13] under the same framework. The Australian Libraries Gateway will not itself be X.500-enabled at this time but the project will test the extent to which migration of the service to the X.500 suite of standards is feasible and the issues that would be involved.

The PRIDE Project has developed a PRIDE synchronization agent that will enable libraries with LIDDAS systems to download directory information from the X.500 test bed server for use in local interlibrary lending and document delivery transactions. A similar synchronisation of directory data is planned between the X.500 test bed server and the national interlibrary lending utility (Kinetica Document Delivery). The intention is to enable this information to be maintained once only through the Australian Libraries Gateway and to be shared between applications engaged in resource sharing activities.

The data modeling stage of this exercise has been completed and work has now begun on development of the new service.

Data modeling outcomes

A key outcome of the data modeling stage has been development of a set of object classes within the PRIDE X.500 directory schema that address the specific requirements of resource sharing directory systems to share data and to automate processes such as filtering of resource providers. In addition the data model for the Australian Libraries Gateway has been re-engineered to provide an overarching directories framework that the National Library of Australia is planning to deploy for all of its directory services, both internal and external.

A core set of entities are defined that enable organizations to be registered and described. This information is shareable between directory services. There is a linking entity that allows information relevant to a specific organizational role such as service administrator to be assigned to a directory collection. Collections currently supported are the Australian libraries Gateway database and the National Directory of Australian Oral History Collections [14]. The new ILRS Directory will form a third collection.

Important issues addressed during the data modeling stage included:

  • The need to revise the existing data model for the Australian Libraries Gateway to support a hierarchical directory structure. (Business rules for parent-child relationships had to be strictly enforced to enable organizations to be identified uniquely by their distinguished name. This required detailed analysis of the other kinds of relationships between organizations that needed to be supported in an unambiguous way.)
  • The need to separate information ubiquitous to an organization from information specific to a given role. (In the model, identity, relationship and type are considered to be properties of an organization. Location is treated as a property of an organizational role, enabling different contact details to be recorded for each role.)
  • The need to enable the ILRS Directory to operate as a directory of national union catalogue symbols. (Identity encompasses the distinguished name of an organization and a range of organizational aliases, including union catalogue symbols. Union catalogue symbols are inherited by collections, locations and services through the organizational role.)
  • The need to resolve ambiguities between collection use policies and access policies. (It was clarified that collection use policies are implicit properties of a collection. They specify whether resources of a given type are lent or copied and under what terms and conditions. Access policies specify classes of user to whom services are provided and under what terms and conditions.)
  • The need to express the complex service parameters and charges informing interlibrary lending and document delivery transactions in a machine-readable form that would support changes in business and service models over time. (A service parameter object class was defined that will enable conditions and charges to be recorded for a flexible range of service parameters, including request, copy, delivery, billing and payment methods, service level, requester location and copyright conditions.)

The draft interlibrary loan directory developed for the North American Interlibrary Loan and Document Delivery (NAILLD) Project [15] provided a valuable cross reference when developing the service parameter object class.

ISO 2146 Project

The modeling work for the PRIDE / ILRS Directory projects also informed the first working draft of the new version of ISO 2146, a standard first published in 1972 to assist in compiling and publishing directories of organizations operating in the library and information domain. This standard is currently undergoing review by ISO TC46/SC4 WG7 [16]. It is planned to issue a committee draft for comment by the end of the year.

The ISO 2146 review has its origins in the standards based modeling approach being practiced by Macquarie University, UKOLN and others. Clearly articulated service models such as the one illustrated in Figure 1 enable definition of the standards and applications to be used across boundaries. Where gaps are identified, duplication of effort and national investment can be minimized and common tools provided by working to develop or enhance standards to address these gaps.

In the case of Australian resource sharing directory services, ISO 2146 sounded like an appropriate standard on which to base application development, but in its current state, it was not suitable. It is tied to a specific syntax. There is no underlying data model. Data element groups are repeated for organizational subunits. Issues such as the difference between collection and service remain unresolved. In addition, substantial revisions are needed to the data elements themselves to take into account the new use cases needing to be supported in a networked information services environment.

The scope and field of application of the existing standard is as follows:

This international standard is intended to assist in compiling directories of libraries, archives, information and documentation centres, including a description of their database services. The rules of this standard concern international directories, national directories published in bi- or multilingual countries and national directories intended for international use and published in one or more foreign languages and national directories. The standard should be used for the collection of the appropriate data and for the publishing of directories as well as for the exchange of data.

The proposed new scope and field of application for the standard is:

This international standard is intended to assist in compiling directories of libraries and related organizations, including a description of their collections (including databases), services and activities. The rules of this standard concern directories operating in an international distributed network environment. The standard should be used for:

  • Collection of the appropriate data.
  • Publishing of international directories, national directories published in bi- or multilingual countries and national directories intended for international use.
  • Exchange of directory data.
  • Provision of access to directory data as part of resource discovery, access management or document delivery processes.
  • Archiving of directory data when the data exists only in electronic form.

Related projects

Resource sharing directory services are just one of the applications to which the ISO 2146 standard will apply. The aim is to build on the directory entities defined to support resource sharing applications by testing the model against a range of other standards and projects. Three such projects are described here.

ISO 8459

The ISO 2146 revision is being conducted in partnership with a project to consolidate the ISO 8459 series of standards into an open-ended object model based on the IFLA FRBR. It is planned to express the semantic and logical relationships between data elements as RDF classes and properties and their syntax and structure as XML structures.

ISO 8459 is a 5-part standard developed by ISO TC46/SC4 WG7 to provide a comprehensive data element directory for library applications. It includes the data elements required to support interloan, acquisitions, information retrieval and circulation applications as well as data elements for the exchange of cataloguing and metadata between systems. Earlier parts of the standard will be reviewed as part of the consolidation process.

Consolidation of the standard should give ISO 8459 greater visibility as a conceptual framework for development of library applications. It is not yet clear whether the ISO 2146 object classes will be an integral component of the ISO 8459 model or a modular extension.

IPIG Directory Schema

The ILL Protocol Implementors' Group (IPIG) is currently engaged in revising the NAILLD draft interlibrary loan directory in the form of an XML schema. An important driver for this project is the need for a directory of ILL Protocol implementors. The IPIG work should contribute significantly to development of the ISO 2146 model in this area. There are similar needs emerging in relation to the data requirements for directories of Z39.50 targets. In both cases, information supporting identification of the organizations themselves should be shareable with other directory services allowing modeling effort to be focused on data requirements unique to the protocol directory service.

Simple Collection Description

Heaney's Analytical Model of Collections and their Catalogues cited above [4] has made an important contribution to the conceptual framework within which ISO 2146 is being revised by treating collections of resources and collections of resource descriptions as sub-types of collection. This was one of the early outputs of the RSLP Collection Description Project [17], which builds on earlier UKOLN work to address the need for machine-readable descriptions of searchable resources.

The RSLP Collection Description Schema bears an interesting similarity to the Z39.50 Holdings Schema, for the good reason that it is addressing the same need to link resource descriptions to directory information. The schema defines a set of attributes "used to describe a collection, its associated location(s) and any associated agent(s)." The data elements used for agents are derived from Heaney's model and parallel work in progress to develop an agent metadata set by the Dublin Core Metadata Initiative [18]. Location is defined as "the physical or online (digital) place where a collection is held". A locator attribute enables direct access to digital locations. Opportunities are provided in free text data elements to record any access restrictions placed on the collection and on the location.

In the Draft Report from the eLib Collection Description Working Group which contains an early draft of the schema [19], data elements are divided into those that describe a collection and those that describe a service providing access to the collection. This more closely reflects the PRIDE Project's distinction between collection and service. There is also more granularity in the data elements provided to record service access conditions.

For the RSLP Collection Description Project a decision must have been made to dumb down the service aspects of the Simple Collection Description. Complex data structures would need to be supported to make this information sufficiently granular to address the PRIDE Project's goal of unmediated access. As "a schema for making relatively simple collection descriptions in a wide variety of contexts -- a 'Dublin Core' for collection description" [20] it is not intended to be a replacement for richer directory schemas where they are needed; for example, as part of a database hosting or aggregation service.

Conclusion

This paper has attempted to describe the directory data standards work currently in progress through the PRIDE, ILRS Directory and ISO 2146 projects and to place this work within a broader context. The need for directory data standards has not been well understood outside the resource sharing community. A driver within that community has been the need to develop sound business models and performance standards for interlibrary loan / document delivery services. There is also a need for resource discovery services that enable the discovery of expressions of works regardless of format and that do not limit the user's options to digital manifestations. This depends on the development of gateway and provider systems that invoke appropriate underlying software and protocols to enable access to resources that are not immediately available online.

The Australian Libraries Gateway experience shows that libraries and related organizations are keen to take responsibility for the maintenance of information about their collections and services when there are efficient processes in place to do so. Over 4,500 libraries now have entries in the Australian Libraries Gateway database. Within this framework it is relatively easy to add new data elements or object classes or to define new directory collections when there is a need to gather new information.

It is still uncertain which protocols will prevail in the delivery of directory services. The X.500 suite of standards offers many advantages, particularly in the context of an organization's own directory service needs. Directory searches can also be supported using Z39.50 or Whois++ protocols. XML query, once it is developed, may prove to be the protocol of choice for many applications because of its ability to treat directory entries as documents and to cope with complex document architectures. For this reason a key aim of the ISO 2146 revision is to express the data structures in forms that are implementation-neutral and that can be easily mapped to other data modeling initiatives.

It is clear that the data structures already developed through the PRIDE and ILRS Directory projects will need to be refined and extended as new functional and mapping requirements are identified. However, linking the ISO 2146 revision to these two projects has provided a strong base for further development to meet the objectives defined in the proposed new scope and field of application for this standard.

Bibliographical references

[1] "Directories - Libraries". In Digital Librarian: a Librarian�s Choice of the Best of the Web. Online. Available: <http://www.digital-librarian.com/dir-libraries.html>. 2000-09-13.

[2] Indecs Home Page. Online. Available: <http://www.indecs.org/>. 2000-09-13.

[3] IFLA. Functional Requirements for Bibliographic Records. Final Report. 1998. Online. Available: <http://www.ifla.org/VII/s13/frbr/frbr.htm>. 2000-09-13.

[4] Heaney, Michael. An Analytical Model of Collections and their Catalogues. Third issue revised. Oxford. 14 January 2000. Online (PDF). Available: <http://www.ukoln.ac.uk/metadata/rslp/model/> <http://www.ukoln.ac.uk/metadata/rslp/model/amcc-v31.pdf> . 2000-09-13.

[5] PRIDE: People and Resource Identification for Distributed Environments. Online. Available: <http://lirn.viscount.org.uk/pride/>. 2000-09-13. The PRIDE definition of Service may be found in the PRIDE Service Data Preparation Guidelines. Online (PDF). Available: <http://www.viscount.org.uk/pride/reppres.html > - Reports <http://www.viscount.org.uk/pride/reppres.html>. 2000-09-13.

[6] National Library of Australia. Register of Australian Archives and Manuscripts (RAAM). Online. Available: <http://www.nla.gov.au/raam/>. 2000-09-13.

[7] Z39.50 Holdings Schema. Version 1.1. July 2000. Online. Available: <http://lcweb.loc.gov/z3950/agency/defns/holdings.html>. 2000-09-13.

[8] McLean, Neil. "The evolution of information resource sharing infrastructure: an Australian perspective". Library Hi Tech. Vol 17(3). 1999. pp. 256-264.

[9] Blinco, Kerry. "Access management for interlending and document supply". Australia/New Zealand Technical Committee IT/19 Seminar. Linking access to electronic library and information services. Melbourne. 21 February 2000 / Wellington. 21 June 2000.

[10] JEDDS (Joint Electronic Document Delivery Software). Online. Available: <http://www.gu.edu.au:81/uls/docdel/jeddshom.htm > <http://www.gu.edu.au:81/uls/docdel/jedsnla3.htm>. 2000-09-13.

[11] LIDDAS (Local Interlending and Document Delivery Administration System). Online. Available: <http://www.slnsw.gov.au/LIDDAS/>. 2000-09-13.

[12] Australian Libraries Gateway. Online. Available: <http://www.nla.gov.au/libraries/>. 2000-09-13.

[13] Australian InterLibrary Resource Sharing Directory. Online. Available: <http://ilrs.nla.gov.au/ilrs/>. 2000-09-13.

[14] Australia's Oral History Collections: a National Directory. Online. Available: <http://www.nla.gov.au/ohdir/>. 2000-09-13.

[15] North American Interlibrary Loan and Document Delivery (NAILLD) Project. Interlibrary Loan directory. Draft. December 1997. Online. Available: <http://www.arl.org/access/naildd/ipig/res/ipig9712-gen-inst-info.shtml>. 2000-09-13.

[16] ISO 2146 Project. Online. Available: <http://www.nla.gov.au/wgroups/ISO2146/>. 2000-09-13.

[17] RSLP Collection Description. Collection Description Schema. V 1.11 2000/05/23. Online. UKOLN. Available: <http://www.ukoln.ac.uk/metadata/rslp/schema/>. 2000-09-13.

[18] Dublin Core Metadata Initiative. Agents Working Group. Online. Available: <http://purl.org/dc/groups/agents.htm>. 2000-09-13.

[19] Simple Collection Description: a Draft Report from the eLib Collection Description Working Group. Online. UKOLN. Available: <http://www.ukoln.ac.uk/metadata/cld/simple/>. 2000-09-13.

[20] RSLP Collection Description. Relationship to ISAD(G) and the Archives Hub. Online. Available: <http://www.ukoln.ac.uk/metadata/rslp/isadg/>. 2000-09-13.

Copyright© 2000 Judith Pearce
spacer
spacer

Top | Contents
Search | Author Index | Title Index | Monthly Issues
Previous Article | Next Article
Home | E-mail the Editor

spacer
spacer

D-Lib Magazine Access Terms and Conditions

DOI: 10.1045/september2000-pearce