A paper written by Judith Pearce (with Janifer Gatenby) and presented in an earlier form at the Standards New Zealand / Standards Australia IT-19 Seminar, Technical Standards for Libraries and Education: Solutions and Emerging Frameworks, held at the National Library of New Zealand, Wellington, New Zealand, Wednesday 26 October 2005.
This paper makes four assertions about the roles of libraries in resource discovery and delivery workflows:
Internet search engines have been seen by libraries both as a threat and an opportunity: a threat in that users may not realise that they are missing important material accessible only through library services; an opportunity as a means of exposing library collections to new audiences through guides, exhibitions and, in recent times, the seeding of search engines with digitised content. Ironically, as more material in library collections becomes discoverable through search engines, their effectiveness as discovery tools is increased and thus also their threat to traditional library services such as the online public access catalogue.
Recently there has been a major shift in the amount of content available or becoming available in digital form. New content is now mostly born digital. Print publications are likely to have a digital master and an increasing amount of information is only being published in digital form. Existing content is also being digitised at an increasingly fast rate. National libraries are engaging in the bulk digitisation of newspapers and full text aggregators are seeking partnerships to do the same for serials. Technologies to capture print materials in digital form, to generate text versions and to analyse their content have enabled digitisation to move from cottage industry status to a high level of production capability.
For books, the Google Print Project and Yahoo's response through the recently set up Open Content Alliance are aiming through different business models to render digital 'the creative output of humankind' (Open Content Alliance, FAQ). Collecting agencies are increasing the rate of digitisation of cultural heritage materials, routinely digitising new materials, copying cover-to-cover and engaging in projects that provide comprehensive coverage in specific areas. In addition, there is no longer any real distinction between creating a copy for interlending purposes and digitising it, at least for out-of-copyright material.
This raises questions at a really fundamental level about the role of libraries in the resource discovery and delivery process and the ongoing need for services such as interlibrary loan. There does seem to be consensus that libraries will continue to build and maintain collections on behalf of their constituencies, managing access to content not freely available online and playing a stewardship role to ensure the survival of culturally significant materials. However, the nature of the resource discovery and delivery services that they deliver and the standards and frameworks needed to support these services will need to undergo a significant change for libraries to interoperate with other stakeholders in this new digital world.
In spite of the rapid growth in digital content identified above, a significant proportion of the material in library collections will continue to exist in physical form. Serials are being digitised at a greater rate than monographs but for books and special materials the situation is different. Dempsey (2005) has estimated that Google Print will digitise around 30% of the titles in OCLC's WorldCat over the next ten years. The volume of physically published material is also continuing to grow and collecting agencies are unlikely ever to have the budgets to digitise all of their heritage material.
It is also important to note that users may still want access for research or learning purposes to a physical copy of resources available in digital form, or to a higher quality copy than is available online. Rights may also apply to digital items that require permission to be obtained before they can be accessed or used. This means that there will be an ongoing requirement for services that support requesting and analogue delivery for at least the next ten years.
This is interesting because a number of recent initiatives to model the resource discovery process do not address this workflow. The E-Learning Framework identifies functions for searching and resolving but not requesting or analogue delivery. Similarly, the NISO MetaSearch Initiative has working groups for collection and service description, search and retrieval and access control but has not included requesting or delivery on its agenda. Delivery is seen as having already been taken care of by OpenURL for online material, a special relationship with the catalogue for resources held locally and NCIP and the ISO ILL Protocol for resources in other collections.
In fact there are still standards issues to be resolved to support seamless discovery and delivery workflows. Seamless in this context does not necessarily mean immediate fulfilment. It means that a user can initiate a process that will result in fulfilment as an integral part of the resource discovery workflow. This requirement needs an end-to-end approach rather than leaving it to separate communities to implement their own service frameworks.
Last year a new version of the ISO ILL Protocol (ISO 10161) was rejected at the Draft International Standard stage because it was not backward compatible with the previous version (Jackson 2005). The new version addressed a number of known problems but more radical changes were needed to justify the development of a new product. The current version uses an outdated technology (ASN / BER1). It is also complex to implement because the protocol is tightly coupled to business rules that would be better handled by applications. A simpler protocol using query response pairs like the NCIP protocol does for consortial lending is now seen as the best way forward, perhaps as an extension to NCIP itself, since there is a lot of overlap between the two protocols. NCIP is an XML and HTTP-based protocol that may itself benefit from updating to the new web service paradigm.
Version 2 of the ILL Protocol has been in existence for eight years but the ISO ILL Maintenance Agency lists just 9 products that support it.. Its uptake has been slow partly because of the complexity of the standard itself and partly because of the many different business models needing to be supported. In Australia, the benefits of the LIDDAS Project (Pearson 2000) are now being seen in a growth in use of the national ILL utility through ISO ILL compliant systems. It will be important to ensure that this impetus continues and that there is a migration path for existing implementations. Resource sharing workflows are inherently complex and there will be an ongoing need for systems able to support these workflows, as well as for simpler interfaces for libraries handling smaller numbers of transactions.
There are still barriers to access, particularly at the point of requesting, that require support for a new set of system interfaces that bridge the current gap between discovery and delivery services, particularly for material in library collections. Users of discovery services are not necessarily presented with delivery options and conditions of access relevant to their context and affiliations. Standards are not in place to enable unmediated requesting or the monitoring of requests that may have been directed to a range of different providers, including commercial providers and online bookshops. For a library service to be competitive with other ways of finding and getting resources on the Internet, the process of requesting a wanted resource needs to be an extension of the discovery workflow for materials of all types.
The Digital Library Federation (2005) has recently identified the need for an abstract service framework for the library community, based on a similar approach to the E-Learning Framework. Business process modelling has been used as an application development tool for many years. A service framework applies this methodology to the business of a whole domain. This enables a community and its stakeholders to get a shared understanding of the common functions needing to be supported to deliver services in a network environment before drilling down to detail.
A useful simplification when talking about the hybrid library collection is to think in terms of 'the wanted resource'. The diagram below shows that physical and digital objects share characteristics such as bibliographic level, whether something is still available from the original publisher or not, whether it is in or out of copyright and whether it is freely available or conditions apply.
Figure 1: The Wanted Resource

A user's capacity to discover wanted resources should not be limited by these characteristics. They will determine which type of delivery service will be used, whether the delivery service needs to support functions such as authentication, authorisation and payment and how the item will be delivered. However the required specialisation only needs to occur at the point where separate workflows take over; for example, when an item is lent, not copied, and therefore has to be returned.
A recent survey by the Research Libraries Group (2005) showed that institutions prefer to target their own catalogues, rather than union catalogues, in their metasearch services. This reflects an economic reality: the need to limit access to union catalogues in order to keep search costs down; and to restrict lending services to library patrons and interlending services to privileged groups within this class. But it also results in a paradox. Libraries enable unmediated access to the world's journal literature through indexes and databases but give priority to their own collections when it comes to the discovery and delivery of books and other non-serial items. Perhaps it is better to charge appropriately for a service as a means of managing costs than to put barriers in place preventing users from discovering the existence of wanted resources.
Union catalogues are still a missing part of the service framework. In order to realise the benefits of the significant investment libraries have made in these tools over the years, they need to be promoted as a primary means of access to wanted resources in library collections. This will highlight the need for standards to support the requesting function as an integral part of discovery workflows and new business models to meet user expectations for timely delivery.
Users are growing increasingly accustomed to self-service, and not just for non-returnable items. Online DVD rental is a good commercial analogue to interlibrary loan. Whether to offer such a service has to be a business decision for each library but systems will need to enable choice by supporting a range of mediated, semi-mediated and unmediated models.
Figure 2 is a snapshot of the Get this item page from the new Libraries Australia service that is now the primary means of access to the Australian union catalogue (Missingham and Boston 2005). It has been designed specifically for end users. Because this screen was accessed from within the National Library, the system knows that the user has an affiliation with the library. The wanted resource is held at the library so the user is provided with a link that takes them through to a record for the same item in the library catalogue, where they can complete the delivery workflow.
Figure 2: Libraries Australia Get this Item Page

If the item had not been available at the National Library the system would have offered the user the choice of requesting (and paying for) a copy or loan or approaching another holding library or trying other suppliers such as second hand bookshops.
In March next year the National Library of Australia is planning to release a free version of Libraries Australia (it is currently a subscription service) and at that time will also start seeding search engines with records from the national bibliographic database in emulation of OCLC's Open WorldCat and RLG's RedLightGreen. This will mean that resources in Australian library collections will appear side by side with free web resources in search engine result sets. When users link to a wanted resource they will be taken directly to the Libraries Australia full record display where, depending on what they system knows about them, they will find a range of delivery options.
Libraries Australia is a web based union catalogue service designed specifically to enable resource sharing. Within the research and higher education library sector the metasearch portal or an e-learning system may be the first point of access to resources for many users and the catalogue or union catalogue may be one of a number of target services. Metasearch services and e-learning systems will need to extend their reference linking services to support requesting workflows for wanted items discovered through union catalogue targets if they want users to have access to resources that are not directly available online.
The full implications of these changes are still unclear but this much is certain: people who have never used libraries before will find material in library collections through Internet search engines; and library patrons will find material in their own collections through a range of different discovery tools and expect the same level of service as if they had found the item through the catalogue. Even now, some users are only using the catalogue user interface to complete the requesting and delivery workflow.
To enable this, the discovery service needs to know how to pass the user through to the record for the wanted resource in the local system. To do this in any scaleable way, libraries will need to implement a resolver service that enables a URL in a standard format to resolve to the syntax used by their catalogue system. OpenURL, already used by full text aggregators for reference linking, is an appropriate standard for this. OCLC is currently developing a resolver service directory to enable this workflow to be fully supported in the Open WorldCat service.
Linking from a union catalogue or portal service to the record in the library catalogue is better than going to the catalogue home page but it can still be an ugly transition. When a library operates both a metasearch service and a catalogue, it is also highly confusing to users to have two competing interfaces to the collection. Ideally the user should be able to complete the workflow from within the chosen discovery service interface. Doing this may well allow some users to bypass the catalogue altogether.
The National Library of Australia is currently looking at what it would take to make this workflow seamless in Libraries Australia for users of its own collection. A number of changes would be required to the library's Integrated Library Management System. The current distinction between lending and interlending and the focus on mediated resource sharing models is reflected in the design of the system. There is an assumption that the requester will be a registered user and this influences the way delivery options are presented in the user interface, the design of request forms and the extent to which users can monitor the status of all their requests in one place. In addition, there is no support for managing the request as a commercial transaction and there is no standard way for a third party document request management system to update the circulation system. The system is also one of many that are still waiting for a firm business case to implement the Z39.50 Holdings Schema.
Figure 3 shows the system interfaces that would need to be in place to enable requests to be placed and managed through a separate discovery service.
Figure 3: Delivery System Interfaces

As a first step, the Request Management System will need to be able to accept requests from a discovery service through a standard protocol and respond with a request number. It will need to accept and store requester details, including payment details, as part of the request without having any prior knowledge of the requester. And it will need to be able to respond to a request for a status report.
Behind the scenes, the Request Management System will need to be able to forward requests to another potential resource provider and monitor the progress of the request. It will also need to be able to update the circulation status of lent resources in the Circulation System. Before the resource is requested, the catalogue will need to be able to respond as a target to requests for holdings information. It should be able to put this information together in the form of a report that includes detailed information not available through the union catalogue, together with the availability of the item and policies related to its use.
These interfaces could be implemented quite easily through a set of simple query / response pairs that targeted the exchange of data needed at various points in the requesting and delivery workflow. Using Application Programming Interfaces (APIs) it would be possible for a union catalogue service to do this on a system by system basis. However, this is not a scalable approach. These are the kinds of transactions that VIEWS (Vendor Initiative for Enabling Web Services) would see as appropriate for its proposed web services architecture should the business need be fully articulated by the library community. VIEWS is an industry wide cooperative effort set up in 2004 to 'make services seamlessly available to the larger world of information handling and processing, whether through tools like Google, through use of e-learning products/services, or via information portals in general' (VIEWS, 2004).
Several communities have already started to support the first scenario using OpenURL or a simple web service for exchanging requests. A web service is probably preferable given the kind of information that might need to be carried. It will also be important for a requesting protocol to fit in with mainstream e-commerce initiatives so that requests can also be lodged with suppliers in other domains and their statuses monitored. Scenarios 2 and 3 are already well supported through the ISO ILL Protocol and the NCIP Protocol, although with the caveats made above. Scenario 4 can be addressed using SRW/ SRU, together with the Z39.50 Holdings Schema for reporting the required information. A new XML Holdings Schema is currently under development by ISO TC46 SC4 that will eventually replace the Z39.50 Holdings Schema for use in a variety of discovery and delivery contexts.
Were these interfaces in place it would be technically possible for the National Library of Australia to decommission the web interface to its own catalogue completely and to use the Libraries Australia interface as the primary means of providing access to its collection. In this scenario the library system would remain a mission critical tool for building and managing collections, contributing resource descriptions to union catalogues, recording detailed holding information and supporting delivery workflows. However, the catalogue would no longer be made accessible as a web based service in its own right, although it would still need to operate as a search target for holdings queries and responses.
For any of the transactions described above to work, the discovery service would need to know which library systems supported the function, the protocol used, the level of support for the protocol and how to access the service. VIEWS has been looking at the need to register web services used by libraries in a central repository of some kind. UDDI is a likely binding for this kind of information. It is an advanced web services discovery standard which has already been deployed in a number of wider business contexts.
The need for registry services to support discovery and delivery workflows extends beyond the registration of web services. The fifth interface identified in Figure 3 illustrates what might be achieved by storing access policies in a central registry. The discovery service itself, based on its knowledge of the user, could query sets of access policies from a range of services in order to make decisions about which delivery options to present to the user. It could also use a registry query to limit the result set to resources that the user is eligible to access or populate resource descriptions with access conditions specific to the user.
Figure 4: Registry Query

Figure 4 is an example of a registry query in which an anonymous user wants an urgent copy of a microfilm. The response contains the decision, charges and conditions. Support for transactions like this will require communities to build resource sharing service registries and make them available through standard search and retrieval or directory protocols. The ISO ILL Protocol Implementers Group (IPIG) has developed an information model for describing resource sharing services both as a means of synchronising information duplicated in local ILL systems and of recording their level of ISO ILL Protocol support. There are no implementations yet, but the National Library of Australia is in the final stages of setting up a test bed using an X.500 version of the schema.
Early findings are that a combination of bindings may be required to support the full information model. X.500 handles the registering of people and organisations and their services but the directory structure is not really suited to the sharing of protocol information or access policies. In the IPIG Schema, these may depend on attributes of the wanted resource, the format in which it is to be delivered, the requesting user, the priority of the request and the requesting, delivery, billing and payment methods used. XACML (eXtensible Access Control Markup Language) may be a more appropriate binding for this part of the model. It is an XML schema for encoding access policies that also includes a protocol for the exchange of access queries and responses.
The metasearch community has also examined the need for registries to describe discovery services. Discovery services are typically based on collections of metadata and / or full text documents. Examples include catalogues, union catalogues, subject gateways, serial indexes, full text databases, institutional repositories, bookshops and Internet search engines. They may be web-based services in their own right or operate as targets in a metasearch service as shown in Figure 5.
Figure 5: Discovery services

There are hundreds of such potential targets and it has become a priority to develop and promote standards that facilitate the sharing of collection and service descriptions to reduce the cataloguing burden. The NISO Metasearch Initiative will be recommending the DCMI Collection Description Application Profile for collection description and the Zee Rex XML Schema for protocol information. This seems like an ideal place for a central registry but the model being discussed at the moment is one where descriptions are shared between metasearch implementers. A central registry would require a business owner and it is not clear which agency or agencies would be willing to take on this role.
The registration model is being used by some communities for their private registries and may become more attractive at a global level if a metasearch service needs to discover targets not already known to it; or if target service providers see value in actively promoting the existence and availability of collections. The nature of the problem may also change as target services are consolidated and the need to authenticate and authorise users in a distributed environment start to be addressed.
Standards for distributed authentication are still under development. The NISO Metasearch Initiative has decided to recommend IP authentication and userids and passwords as access control mechanisms for the time being but sees Shibboleth as a promising distributed authentication method and certainly one to watch. It allows users requesting access to a service to be authenticated by another identity provider such as their university. A Shibboleth federation may require the registration of federation members and services in a central registry. There will also be a requirement to register the authorisation attributes required by service providers in order to deliver a specific level of access.
ISO 2146 is an international standard currently under development by ISO TC46 SC4 to operate as a framework for building registry services for libraries and related organisations. Its current title is 'Registry Services for Libraries and Related Organisations'. It takes the form of an information model that identifies the objects and data elements needed for the collaborative construction of registries of all types. It is not bound to any specific protocol or data schema. The aim is to be as abstract as possible, in order to facilitate a shared understanding of the common processes involved, across multiple communities of practice.
The standard has been under development for several years now. An earlier draft was based on the IPIG Schema and still had many data elements relevant only to resource sharing services. This has now been fully generalised and an advanced draft will be discussed at the ISO TC46 SC4 Meeting in Thailand in February2 2005.
Figure 6 shows that a registry service can have as its primary registry object any one of four sub-classes relevant to the library community: collection, party, activity or service. This allows it to have more general application than the IPIG Schema, which requires the primary registry object to be an organisation. As well as supporting the description of resource sharing services, ISO 2146 could be used as the framework for a registry of metasearch collections and services through the Collection object or a registry of training courses or projects through the Activity object.
Figure 6: The Registry Object

A party can be either a person or a group of people, for example, a family or organisation. Parties collect or own collections and organise or participate in activities. Parties also manage and use services through which collections and activities are accessed. Services may have policies that permit or deny access to users under certain conditions. In a network environment, services may be delivered through protocols. A registry is itself a collection that can be accessed through a discovery service using standard protocols.
Service is the most complex of the four sub-classes of registry object. A service of a given type will functions, products and users and may have access policies if it is not freely accessible to all users. An access policy may permit or deny access based on the function being performed, the product being requested and the user making the access request. In Figure 7 these make up the target of the access policy as defined in the XACML schema.
When access is permitted there may be a range of obligations such as a requirement to pay a charge or a specific loan period and overdue charges. These may be dependent on service level. For each service level supported, service descriptions need to be set up in a machine-parsable form. In Figure 7 these make up the obligation of the access policy as defined in the XACML schema.
Figure 7: Service

Identifying the services needing to be registered by a community in order to do business in a network environment is closely allied to the development of a service framework. In fact, the two go hand in hand. The high level functions in a service framework should be exactly those functions that need to be enumerated in access policies and associated with protocol information in service registries.
The functions listed in Figure 8 for use as the default enumeration for the Function element in the ISO 2146 Service object can also be seen as a draft service framework for discovery and delivery services. These include functions that are common across services (Authenticate, Authorise, Pay); functions specific to metadata services (Contribute, Save, Alert, Harvest), functions specific to discovery services (Find, Locate, Request), functions specific to delivery services (Resolve, Supply, Lend, Reserve) and functions needed to personalise the user's interaction with the service (Register, Ask, Personalise, Monitor). Registry Services may have a role in the execution of any of these functions as well as being discovery services in their own right.
Figure 8: Functions

Including Request as a discovery function in this model may seem a small thing but it reflects a major change in thinking about delivery services, as does combining lending and interlending into a single function and the supply of non-returnable copies with the supply of other purchasable products.
This paper has argued that the future of the online public catalogue is uncertain and presented the case for the union catalogue as a primary target for access to library collections, either as a web-based service in its own right or a metasearch target. However, the extent of change in the discovery framework will not stop there.
Figure 9: Metadata

The diagram in Figure 9 models the metadata needed to facilitate the discovery and delivery of wanted resources. The resource description itself, content listings and holdings are familiar categories of metadata but we do not usually consider full text indexes as metadata.
Until now there has been a relatively clean division between discovery services based on metadata collections and those based on full text databases. As more material becomes available in digital form, users will expect discovery services to be able to search the full text of these resources in conjunction with any associated metadata, even when the resource itself is stored in a separate repository.
To meet this need discovery services will have to become search engines, capable of indexing and ranking distributed content. Metasearch services will also need to change. The current metasearch model is agnostic to the number of collections being searched. This imposes on the user the burden of knowing which collections to select and also results in duplication in result sets and the dumbing down of semantics inherent in the metadata. In addition, while individual targets may rank results the metasearch service has no capability of interpreting the rankings in the context of the whole corpus of information being searched. Lorcan Dempsey (2005a) is now increasingly wondering 'if it might make more sense to attack the boundary issue differently, by working on business and technical approaches which would result in fewer, larger resources to search'.
The National Library of Australia has started to put this philosophy into practice, at least with its own targets. The metadata collection that is used to deliver the Music Australia service is derived from the Australian National Bibliographic Database (ANBD) that supports the Libraries Australia service. A metasearch service targeting the ANBD can trust that it does not also need to target the Music Australia collection to find Australian music resources. However, another web site only interested in music could target just the Music Australia collection.
This approach to the metasearch problem is proving its value and the Library is looking at forming similar trusted aggregations for research materials in institutional repositories, for pictures and for the corpus of Australian journal material. Initiatives such as Google Scholar and the bulk digitisation of serials are likely to have an impact on the number of target services needed to for comprehensive coverage of the journal literature and provide opportunities for rationalisation and consolidation of indexing and full-text aggregation services. Museums and archives are also increasingly looking to federated services and resource sharing models to enable the kind of discovery and linking workflows needed to bring related material together.
The trend towards larger aggregations and the competition from services like Google Scholar is likely to change the nature of commercial services as well. There is a growing expectation for the discovery interface to be free and an acceptance that delivery may be subject to conditions. New business models will be needed to enable metadata and full text indexes to be aggregated and deployed in a range of different 'online spaces' while also enabling the authentication, authorisation or payment workflows needed to support delivery.
The a9.com website which went live in April 2004 is an example of a completely unconstrained metasearch model. The protocol used is OpenSearch. Based on RSS, it works as a simple query, allowing targets to interpret the search terms and present the results that they think are most relevant. Targets register themselves. Users select from a default set, or add more targets from a pull-down list. A registered user can specify their favourite targets. Results are displayed side-by-side as separate result sets.
There is a lot to be learnt from this model: the inclusion of Google, the Google image search, amazon.com book search and the Wikipedia amongst the default targets generally gives good results, taking into account the coverage of the content being searched. The capability to personalise the user interface lets users find their own trusted aggregations and add them to the defaults. Once registered they become capable of engaging in e-commerce transactions. The dependence on the target to make the decisions about what is relevant and the use of RSS to present results frees the client software from any need to support fielded searching or have an undertanding of underlying indexes or data schemas.
Discovery services providing access to library collections will need to compete with this level of usability and ease of implementation. The challenge will be to meet users' expectations that they can find wanted resources through a simple keyword search while optimising both precision and recall. One way of meeting this challenge will be to build intelligent clients that understand the context for the search, who the user is and the nature of the target and can frame the search appropriately. An intelligent client might have a special relationship with a trusted aggregation such as a union catalogue, allowing it to make its own decisions about what records to return. It might also hide the process of selecting the appropriate targets to search, guiding users to trusted aggregations or filters through terms such as 'books', 'articles', 'pictures', 'papers' etc. It might deploy these choices in a whole range of different contexts, for example, as a follow-up search on a page in an encyclopedia or dictionary of biography or gazetteer.
Most mainstream search engine products have no support for standard search and retrieval protocols. Their business model is based on distributed indexing and centralised searching. They all have standard application programming interfaces that can be used to deploy the search engine as a target. This reflects the fact that the enterprise is the primary market for search engine vendors rather than that there is no longer a need for standard search and retrieval protocols. The business case for an abstract protocol in the metasearch environment has not changed at all, as the development of OpenSearch shows. There is also still an ongoing need for a protocol that supports a query language and fielded searching. The more effective the user interface, the more use an intelligent client may be making of these features in the construction of the search before sending it to the target.
In the library and information domain, SRW/SRU is still the preferred standard for this. Its main competitive advantage compared to OpenSearch and other search protocols is CQL, its query language and its association with Z39.50 through legacy implementations. While implementing a query language is not technically complex, it requires a community to have an agreed understanding of the semantics behind the language. Some search clients and targets may never have a sufficient business case to implement SRW / SRU. For this reason, library systems and services will certainly find themselves needing to support more than one search protocol.
When extending the boundaries of a metasearch service to new communities of practice such as museums and archives, it will be important to identify the exact workflows that really need fielded searching. It may be that, instead of trying to address the semantic interoperability issues inherent in implementing a query language, the effort could better be spent building trusted aggregations that can behave like search engines, delivering ranked result sets based on algorithms appropriate to the corpus, with fast performance.
This is another reason for having fewer, larger targets. Centralised systems and large corpuses of information lend themselves better to the kind of analysis that is needed to provide effective ranking algorithms and filters. The National Library of Australia plans to use FRBR to cluster related resources in the ANBD and to experiment with new technologies such as topic maps to exploit semantic relationships inherent in (or deducible from) the content being indexed. This effort might then be shared with other services either directly as a search target; or through the harvesting of metadata to form other collections.
It is important to note that this is not implying that there will no ongoing need for metadata or controlled vocabularies. For data not yet online, information in the existing resource description will be used to rank the resource. For digital material, there will be more information available to analyse but it will not be controlled. Controlled vocabularies will still play an important role in bringing like material together and there will be a benefit in storing the results of data analysis and mining in resource descriptions for resource sharing purposes. However, the whole process of cataloguing is likely to be changed, with systems automatically generating metadata as part of the indexing process.
There is no longer any real boundary between library systems and the deep web and Internet search engines. Partnerships and alliances are emerging that are giving users a level of access to resources through search engines that we could not have imagined even five years ago. In this context libraries still have a role to play but it will require the implementation of new business models and service frameworks, both for resource discovery and delivery. As union catalogues are transformed from tools for librarians to end user tools, the library system will lose its primacy as a discovery service but need to take on new roles to support discovery in a distributed environment. The discovery interface will be free via search engines or a metasearch service. Infrastructures will be in place that support the requesting and delivery process and any associated authentication, authorisation and payment workflows in a seamless fashion. Libraries will still use their catalogues and local request management and circulations systems to manage this process. Standards will continue to play an important role, but not in the form of systemic protocols. To respond quickly to new requirements, systems will need to be modular, with the capability of engaging in a transaction with another system through simple query / response pairs. Registries will also play an important role, providing the information needed to enable systems and services to work together to support end-to-end workflows.
Childress, Eric (2005) Pattern recognition for technical services: Interpreting
the OCLC Environmental Scan, Ohio Library Council annual conference &
exposition, 6 October 2005, Columbus, Ohio (USA).
http://www.oclc.org/research/presentations/childress/20051006-olc.ppt
Dempsey, Lorcan (2005) 'The inside out of libraries: libraries in the age
of Amazoogle' LIBER pre-conference, Groningen, 5 July 2005
http://liber.ub.rug.nl/presentations/Dempsey.ppt
Dempsey, Lorcan (2005A) 'Metasearch: a boundary case' Lorcan Dempsey's weblog
on libraries, services and networks
http://orweblog.oclc.org/archives/000768.html
Digital Library Federation (2005) DLF service framework for digital libraries:
a progress report for the DLF Steering Committee 17 May 2005
http://www.diglib.org/architectures/serviceframe/dlfserviceframe1.htm
Jackson, Mary (2005) 'When a good standards process fails' Interlending and document supply v33, no. 1 2005, pp. 53-55
Middleton, Roxanne and Boston, Tony (2005) 'Libraries Australia : Creating a new national resource discovery
service' An article for Online Information Review also published as
a staff paper on the Tational Library of Australia website
http://www.nla.gov.au/nla/staffpaper/2005/missingham4.html
Lynch, Clifford (2005) 'Where do we go from here? The next decade for digital libraries', D-Lib Magazine, Volume 11, Number 7/8, July/August 2005 http://www.dlib.org/dlib/july05/lynch/07lynch.html.
Open Content Alliance
http://www.opencontentalliance.org/
Pearson, Kathryn 'Changing the face of interlibrary loans: LIDDAS at Macquarrie
University' VALA Biennial Conference and Exhibition (10th: Melbourne:
2000)
Books and bytes : technologies for the hybrid library
http://www.vala.org.au/vala2000/2000pdf/Pearson.PDF
RLG (2005) Metasearch survey among RLG members.
Survey report
http://www.rlg.org/en/page.php?Page_ID=20750
Shenton, Helen (2005) 'Strategic developments in collection storage' LIBER
conference, Groningen 6 July 2005
http://liber.ub.rug.nl/presentations/Shenton.ppt
VTLS (2004) Views: Vendor Initiative for Enabling Web Services
http://www.views-consortia.org
Google Scholar Beta (2005) About Google Scholar
http://scholar.google.com/
Google Print Beta (2005) About Google Print
http://print.google.com/intl/en/googleprint/about.html
National Library of Australia, Libraries Australia: the
Kinetica Search Service
http://librariesaustralia.nla.gov.au/apps/kss
National Library of Australia, Music Australia
http://musicaustralia.org
OCLC Open WorldCat
http://www.oclc.org/worldcat/open/default.htm
RLG (2005)RedLightGreen
http://www.redlightgreen.com/ucwprod/web/workspace.jsp
The E-Learning Framework
http://www.elframework.org/
Interlibrary Loan Application Standards Maintenance Agency
http://www.lac-bac.gc.ca/iso/ill/
Internet 2 Shibboleth
http://shibboleth.internet2.edu/
ISO Technical Committee 46 Subcommittee 4 Working Group 10
XML Schema for Holdings Information Project
http://www.unt.edu/xmlholdings/ScopeOfProject.html
NISO Circulation interchange Protocol (NCIP). Standards Committee AT
http://www.niso.org/committees/committee_at.html
NISO NISO Metasearch Initiative
http://www.niso.org/committees/MS_initiative.html
NISO OpenURL framework for context-sensitive services: Standards Committee
AX
http://www.niso.org/committees/committee_ax.html
OASIS eXtensible Access Control Markup Language (XACML) TC
http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=xacml
OpenSearch
http://opensearch.a9.com/
UDDI
http://www.uddi.org/
Z39.50 Maintenance Agency Z39.50 Holdings Schema
http://www.loc.gov/z3950/agency/defns/holdings.html
Z39.50 Maintenance Agency SRW - search/retrieve web service
http://www.loc.gov/z3950/agency/zing/srw/
| About This Site | Copyright | Privacy | Accessibility | Site Map | Site Search | Content A-Z | Contact Us |