PECAN Final Report

51
FINAL REPORT PECAN Pilot for Ensuring Continuity of Access via NESLi2 EDINA and JISC Collections

description

Ensuring continuity of access to e-journal content is a key concern to libraries, in terms of both the need for long-term preservation and post-cancellation access. The PECAN project is investigating the policy, procedures and infrastructure needed to provide our members with guaranteed post-cancellation access to their NESLi2 e-journal holdings. This is the report from the project

Transcript of PECAN Final Report

Page 1: PECAN Final Report

FINAL REPORT

PECAN Pilot for Ensuring Continuity of Access via NESLi2 EDINA and JISC Collections

Page 2: PECAN Final Report

Index

Overview of Project 1. Introduction 32. Aims and Objectives 43. Overall Approach 4

Section One . Investigation of the issues of post cancellation access for UK HEIs and the use of ‘trigger events’ and appropriate governance arrangements for PECAN 6Supply of Post-Cancellation Access via PECAN: A Back-Up Mechanism 6 Cancellation as Trigger Event 7Governance Arrangements 7Proposed Entitlement Registry: Data Sources 7Proposed Locate Registry: Data Sources 8Proposed Central Archive: Data Sources 9Use Cases 9

Section Two Survey of librarians and publishers attitudes to Registry of Entitlement and Secure Archive 11Findings 12Comparison of subscription data between publishers and librarians 15Findings from the comparison 16Section conclusions 16

Section Three

Candidate technical infrastructure for a registry of entitlement and a secure archive 18Functional Overview 19Overview of Components 19

Conclusions and Recommendations

Conclusions 25Recommendations 26

APPENDIX A - An Exploration of Issues 27

APPENDIX B - List of all additional comments from libraries responding to online Survey 32

2

Page 3: PECAN Final Report

Proposal for JISC funded Project

Pilot for Ensuring Continuity of Access via NESLi2 (PECAN)Overview of Project

1. Introduction

The assurance of continuity of access to scholarly material in the form of e-journals is an important matter for all academic and research libraries. The concern is for long-term preservation and also for access to back content in the event of cancellation of current subscription. This concern is at all levels in UK higher education: Higher Education Institutions (HEIs) and the bodies that fund them. Additional measures are required to protect the investments that organisations have made when procuring electronic materials that support teaching, learning and research.

The focus of the scoping study funded as the PECAN project has been on post-cancellation access, which has proved particularly significant and timely with the onset of budget restrictions. This form of access to online back content has not been generally addressed by libraries and publishers. Of the organisations set up to provide e-journal archiving services, such as CLOCKSS, Portico etc, their focus has been geared towards digital preservation. An exception to this is material stored as part of the UK LOCKSS (Lots of Copies Keep Stuff Safe) Alliance networks. However, UK LOCKSS has had to gear itself towards journal titles not owned by the larger publishing organisations that have preferred to participate in CLOCKSS and Portico. Many of those larger publishers have robust infrastructures of their own in place, and are currently providing access to materials post-cancellation through these infrastructures.

The NESLi21 scheme was devised by the JISC and JISC Collections to assist UK HEIs with the procurement of electronic journals. It is proposed that this framework be extended to provide the basis for more robust post-cancellation access arrangements between publishers and consumers of e-journal material. Three facilities are envisaged: a registry of entitlement (which has reliable information on the journal content that has been subscribed by libraries via NESLi2); a locate registry (re-directing requests to points of service delivery on that content); and a secure (central or virtual) archive providing assured continuity of access to back journal content.

Arrangements are in place between archiving agencies and publishers to lift restrictive licensing arrangements in the event of a number of defined ‘trigger events2’ and this is 1 NESLi2 is a HEFCE funded national initiative for licensing of electronic journals on behalf of higher and further education and research communities. NESLi2 negotiates with the major scholarly publishers, and of these, the majority deposit content in CLOCKSS.2 Trigger events, presupposing no successor interests or reversion or transfer of rights, include (1) publisher of content in question no longer in business, in business of publishing content, or providing access to this content; (2) publisher of the impacted content has stopped publishing, is no longer providing access to the content; (3) publisher has stopped offering or providing access to some or all of back issues of triggered content; (4) While still publishing the content, publisher is not able to provide access to the content electronically due to technical or similar catastrophic and permanent failure.

3

Page 4: PECAN Final Report

addressed as part of the NESLi2 negotiated agreements. Those trigger events do not allow for access if a library has to cancel its subscriptions to a journal or collections of journals. The NESLi2 Model Licence does contain a post cancellation clause, and what is important now is that methods are found to support the use of that clause. This is especially important in the current economic climate, when many libraries’ budgets are reduced due currency fluctuation impact and reductions in budgets.The intention is to investigate how to extend the functionality achieved by UK LOCKSS Alliance, CLOCKSS (Controlled LOCKSS)3 and similar ‘private LOCKSS network’ in order to ensure continuity of access, post-cancellation.

2. Aims and Objectives

The project aimed to investigate how post-cancellation access for NESLi2 deals can be achieved by reaching agreements with publishers, so that libraries can access material stored in the CLOCKSS archive or a similar ‘private LOCKSS network’.

The objectives were:i. To fully examine the complexities of defining what might constitute institutional

entitlement to post-cancellation access and to scope a network-accessible registry.

ii. To examine and propose potential governance arrangements for administering and monitoring post-cancellation access

iii. To examine and propose a candidate technical infrastructure that would provide appropriate controlled access to licensed material within the CLOCKSS archive, in a robust and secure manner.

iv. To consult with publishers and other stakeholders to establish whether the proposed measures could form part of the NESLi2 licensing deal.

3. Overall Approach

An outline of the methodological approach is set out in Appendix A. Two facilities were envisaged: a registry of entitlement; and a secure archive. The first would be a network-accessible facility containing reliable information on the journal content to which libraries have subscribed via NESLi2. This would be populated in two stages: as assertions from libraries; and with confirmation from the relevant publisher. The second was the means to provide secure and robust access to back journal content, being either an actual archive or assured means of network access to trusted archives of content.

The methodology assumed that a significant amount of formal consultation and negotiation with the seven individual NESLi2 publishers would be involved.

3 CLOCKSS is a joint venture between the world’s leading scholarly publishers and research libraries whose mission is to build a sustainable, geographically distributed dark archive with which to ensure the long-term survival of Web-based scholarly publications for the benefit of the greater global research community. Publishers are joining CLOCKSS to preserve their titles in a secure dark archive without the cost-prohibitive expense of building their own preservation and storage system. It is for the CLOCKSS Board to decide when a publisher's content is triggered. CLOCKSS should not charge libraries for access; instead, the triggered content is made freely and globally available from designated CLOCKSS host sites, currently Stanford University Libraries and EDINA, University of Edinburgh.

4

Page 5: PECAN Final Report

The final report covers the following:

1. Investigation of the issues of post-cancellation access for UK HEIs, and the use of ‘trigger events’ and appropriate governance arrangements for PECAN

2. Survey of Library and Publisher attitudes towards issues around post-cancellation access and the establishment of a registry of entitlement and central archive facility.

3. Candidate technical infrastructure for a registry of entitlement and a secure archive

5

Page 6: PECAN Final Report

Section One

Investigation of the issues of post-cancellation access for UK HEIs, and the use of ‘trigger events’ and appropriate governance arrangements for PECAN.

Clause 8.5 of the NESLi2 licence provides the contractual basis for continued access post-cancellation but implementing this is new and unproven ground. Two questions arise: to what content is an institution entitled to access and how can continued access in perpetuity be guaranteed?

One element of the PECAN study has been to survey a sample of NESLi2 librarians and publishers to compare understanding about which journals would be covered within the NESLi2 deals, and to gauge reaction to an outline information architecture.

A consultation event was held at the JISC offices in London in early October 2009 to review the summaries being reported. There was support for an entitlement registry - a subscriptions database with history - to assist libraries. The idea of a central archiving facility was also discussed. There was shared priority across the publishers and librarians present that researchers and students should have as much ease of access to back copy post-cancellation as to currently subscribed content. The challenge was to devise and provide a location facility that would resolve access to the appropriate, and assured, source of supply.

We have identified four major components that comprise our proposed information architecture.

1. A post-cancellation resolver service that responds to requests from library patrons and third party serials services, via both a browser service and m2m interoperability, by utilising entitlement and locate registries,

2. An entitlement registry that matches up title information with institutional subscriptions and post-cancellation entitlement,

3. A locate registry to redirect the enquiry towards potential and appropriate sources of supply, and

4. A central archive of content that provides assured continuity of access.

Supply of Post-Cancellation Access via PECAN: A Back-Up Mechanism

The objective of PECAN is to improve availability of accurate historical subscription information and access post-cancellation. PECAN is not driven by an intention to replace or disrupt the current models of post-cancellation access in the publisher supply chain, and we recognise that supply of post-cancellation access must take into account the requirements of both libraries and publishers. Libraries wish to have assurances that they will receive access to content to which they are entitled, and the source of supply is arguably of less concern. Increasingly, publishers are supplying post-cancellation access themselves, and correspondingly the conditions of use associated with any national or third party archives require a degree of negotiation to operate alongside publisher priorities. In PECAN, we are proposing as one component a central archive to provide assurance of supply to libraries, however this is not intended to replace the publisher as the main source of supply. This has been taken into account via the proposed locate registry, which is intended to allow a publisher to specify a preferred source of supply.

6

Page 7: PECAN Final Report

Cancellation as Trigger Event

The entitlement registry will comprise a database of current subscription to journal content, with history. This history will record changes in subscription status. For example, cancellation of subscription is an event to be recorded. This change in subscription status may be used by service suppliers to authorise post-cancellation access to journal content. In the proposed central archive, access to content shall only be authorised when permitted by the subscription licensing conditions recorded within the entitlement registry.

Several e-journal dark-archiving services have emerged over recent years. These provide archives of online content that can only be accessed by users when strict conditions (known as ‘trigger events’) are met. These trigger events tend to be associated with orphan content; that is, content not available from any other source. PECAN differs from this model in that access is necessarily associated with entitlement as expressed through license.

Governance ArrangementsTo ensure the continued support of publishers and other interested parties (e.g. libraries and subscription agents), there must be opportunity and mechanism for them to have sufficient input on the governance of the PECAN service. The responsibility of this group will be to consider and develop the role, scope and policies surrounding the PECAN post-cancellation resolver service and its entitlement and locate registries, alongside existing sources such as publishers and third-party services (e.g. CrossRef, http://www.crossref.org). A strong governance model is essential in maintaining stakeholder trust in the service. This trust includes ensuring the security of the registry contents, agreement over the trigger mechanisms for release of content to entitled institutions post-cancellation and maintaining appropriate quality of the data held. For the service to be useful and reliable it ought to have longevity. While project funding can help establish a service, ongoing funding models and/or a sustainable business model should be considered from the outset.

The registry of entitlement should also consider issues of confidentiality surrounding the management of information on library holdings and the availability of such information to the different stakeholders, given the potential commercial sensitivities involved.

Defining in detail a model of governance is outside the scope of this report, as it requires extensive discussion with all stakeholders involved, namely, the central funding agency, JISC Collections, publishers and librarians

The PECAN central archive should be considered a separate service from the entitlement registry and can be governed by JISC Collections / National Data Centre. The important thing is that publishers agree to the terms of access. This really comes down to the Locate Registry – the preferred routes that are set and the mechanics of linking.

Proposed Entitlement Registry: Data SourcesThe entitlement registry shall hold authoritative and trusted records used to grant institutional access to subscription content, and so the accuracy and up-to-datedness of this data is critical. A reasonably comprehensive level of participation from publishers is

7

Page 8: PECAN Final Report

required or the registry will suffer from poor data quality and low usage. It is likely that JISC Collections shall be essential in encouraging participation, perhaps as a requirement of NESLi2. 

The support of publishers and/or subscription agents is needed, and there must be mechanisms for publishers to inspect, verify and correct data. An audit process must be in place to demonstrate that data is sufficiently up-to-date, and that the registry contents are accurate. This is critical; the registry would be undermined by poor information quality and a lack of timeliness. As much as is reasonable, data flows, timings of updates and data formats should build upon mechanisms currently in use.

It is likely that updating the entitlement registry will be an additional step in the workflow between a subscription agent and publisher. When a library orders a subscription (generally via an agent than direct with a publisher), the library expects this subscription to continue unless they inform the agent (or publisher) that they want to cancel (and they need to do this in time, before the agent sends renewals to publishers). Publishers only enter renewals in their systems based on the data they receive from agents annually. To inform publishers, subscription agents are using electronic data interchange formats such as ICEDIS (http://www.icedis.org/).

Many publishers now provide post-cancellation access to 'unsubscribed' titles (i.e. the titles in Big Deals that are not 'core subscriptions'). This means that a NESLi2 deal with a publisher does not just refer to the core set of subscribed titles but to a broader package of titles, with some core subscriptions and some unsubscribed titles. The extent to which subscription and entitlement data needs to be collected requires consideration. There are two options: the entitlement registry can record every subscription, specifying all titles in a Big Deal package and the dates for which post-cancellation access is authorised; or the registry can add data only at the point of cancellation for the set of cancelled titles.

For use in an entitlement registry, further analysis is required to identify the most efficient, clear and comprehensive method by which entitlement can be recorded, based on how subscription information is conveyed between subscription agents and publishers. The process for recording termination and renewal is central to the successful operation of an entitlement registry. Equally important is the process used for recording changes to the set of titles for which post-cancellation access is permitted. Maintaining this information requires the support of the actors involved in updating this information, namely publishers and their subscription agents.

Proposed Locate Registry: Data SourcesA locate registry is proposed to direct a request for access towards potential and appropriate sources of supply. The contents of the locate registry can be considered as somewhat analogous to link resolver knowledge bases (KBs). Further analysis is required to consider the extent to which submission processes for publishers into link resolver KBs can be leveraged for the submission of location information. It will be critical that the information submitted is actionable; OpenURLs shall be used to identify and link to appropriate resources.

The extent to which location information is readily available from alternative sources of supply, such as dark archiving initiatives, is not yet known. If this data is unavailable, this may be because there has not yet been high demand for location information. To date,

8

Page 9: PECAN Final Report

link resolver systems have focused on current subscription, though this may change as the quantity of orphaned titles increases. Dark archiving initiatives are new services, and the significant focus of their activity to date has been on encouraging publisher participation and collecting content. Introduction of a locate registry dedicated to post-cancellation access may stimulate best practises and contribute towards wider accessibility of location information.

Proposed Central Archive: Data SourcesA central archive for the UK library community should in the first instance, comprise material from publishers participating in NESLi2. The involvement and support of JISC Collections can be expected to encourage publisher participation, however publishers are known to be sensitive about permitting access through alternative sources of supply, and thus the terms and conditions under which content shall be accessed will need to be explicitly negotiated with publishers. Clarity is needed on how much responsibility publishers are willing to take for provision of post-cancellation access. Policy makers in the UK need to assess whether current models of supply (via publisher and/or dark archive initiatives) are sufficient, or whether further assurance of access is needed by locating content within UK-based systems and infrastructure, and operated under local policy control.

Building upon existing software may also permit reuse and extension of existing publisher workflow processes. For example, seven of seventeen NESLi2 publishers are already participating in CLOCKSS (http://www.clockss.org), a dark archive of electronic scholarly content jointly governed by publishers and librarians on behalf of the global community. If a central archive were to be established in the UK, it may be attractive for publishers to use technologies and workflow processes already in place.

Use Cases1. Following cancellation of a journal title, a library patron (researcher or student)

uses the PECAN service to find an appropriate source of supply: a. A library patron requests a journal title through a library catalogue that

has been configured to direct requests for access to the PECAN resolver service.

b. The PECAN resolver service receives the request and identifies the institution making the request, and refers to the entitlement registry to authorise access.

c. If successful, the PECAN resolver service then refers to the locate registry to identify the preferred and appropriate source of supply.

d. The PECAN resolver service then forwards the library request to the appropriate source of supply, along with an authorisation token.

2. A librarian uses the entitlement registry to clarify and confirm subscription status and post-cancellation access entitlement:

a. When a library subscribes to content, a record of that subscription is entered into the entitlement registry either by the publisher or a subscription agent acting on their behalf.

b. The library and publisher can then refer to the entitlement registry at any point to establish comprehensively their subscription holdings and post-cancellation entitlement for particular publishers.

9

Page 10: PECAN Final Report

3. Publisher uses the PECAN locate registry to indicate the preferred supplier of post-cancellation access to content:

a. E.g. the default supplier is the publisher’s own copy.b. If that is unavailable, PECAN defaults to a UK HE central archive (if one

were to exist) unless the publisher indicates an alternative.c. The default does not exclude the availability of other sources of supply. A

policy decision is needed on whether seamless linking to the preferred source is implemented, or whether it is more appropriate to present to the user a click-through set of source options.

Additional Use Cases4. Librarian uses both the entitlement registry and locate registry to resolve the

matter of (post-cancellation) access following transfer of title ownership between publishers. Resolving entitlement to back-copy content following transfer is a necessary activity irrespective of whether the focus is post-cancellation access or not. Standard procedures may not yet necessarily be widespread, and PECAN may be able to assist the process.

a. Following title transfer, a new publisher takes ownership of the relevant entries in the entitlement registry.

b. The default source of supply in the locate registry is updated to point towards the new publisher. If the new publisher chooses not to provide post-cancellation access, the locate registry defaults to the central archive.

10

Page 11: PECAN Final Report

Section Two

Survey of librarians and publishers attitudes to Registry of Entitlement and Secure Archive

Interviews were held with a small number of NESLi2 publishers and librarians to inform them of the initiative and obtain their feedback and views on the proposed Registry and secure archive. In these discussions we also sought their agreement to participate in an exchange of journals subscription data between publishers and librarians to compare understandings of which content would continue to be available to libraries after cancellation of subscriptions or termination of an agreement with publishers. We also invited them to a meeting with JISC Collections, EDINA and CCL where additional information on the initiative would be presented and their feedback sought

These interviews were conducted by a mixture of telephone and face to face meetings.

The discussions with librarians and publishers also included reference to the inclusion in the NESLi2 Model Licence of a key clause referring to post termination access, to illustrate that publishers would already be aware of the potential requirement by JISC Collections to obtain publishers’ data for ‘local loading’ within the UKHE community.

– After termination of this Agreement […] the Publisher will provide […] the Licensee […] with access to and use of the full text of the Licensed Material which was published and paid for within the Subscription Period,

byi) continuing online access on the Publisher's interface,ii) a central archiving facility operated on behalf of the UK HE/FE

community, or – or other archival facility

iii) supplying the Licensee with archival copies

Publishers and librarians interviewed were:

– Elsevier– Institute of Physics Publishing– Oxford University Press– Royal Society of Chemistry– SAGE– Springer– Wiley-Blackwell

– University of Aberdeen– University of Cardiff– Edge Hill University– Glasgow University– University of Hertfordshire– UCL

11

Page 12: PECAN Final Report

Additionally an online survey was undertaken amongst NESLi2 institutions. Questions included in this brief survey were:

1. Would you welcome the creation of a Web-based registry recording the access archival rights relating to NESLi2 journals, for each institution participating in NESLi2 agreements? (It is anticipated that the registry would list individual journal titles and individual institutions, showing the periods that each institution is entitled to post termination access to each title)

2. What do you think would be the benefits of such a registry?3. What do you think would be the challenges / difficulties (if any) in creating and

maintaining such a registry?4. Would you welcome the development of a central dark archive of NESLi2 e-

journals within the UK HE community? 5. It is anticipated that for individual titles that have been cancelled, access by

institutions would continue to be obtained from individual publishers’ websites. The archive would come into use in situations where events have caused access from publishers’ sites to be unavailable. Would you agree with this approach?

6. What do you think would be the challenges / difficulties (if any) in creating and maintaining such an archive?

7. Any other general comments you may wish to provide:

FINDINGS:

A. from the interviews:

Librarians:

Libraries welcome the initiative and are willing to compare data with publishers The registry could potentially be of more interest than the archive Creating and running an archive could be a massive project Potential problems with title changes, transfers, mergers Challenges in reaching agreement with publishers on rights The registry would be very useful for reassurance of actual rights It would give comfort to libraries moving to online-only How does this affect LOCKSS and Portico? Keeping the registry up to date would be a major task How would the archive and registry be funded? What would be the ‘start’ date of content in the archive? Would the archive eventually include non-NESLi2 titles? ERMs could help in provision of lists of eligible titles

Publishers:

Publishers broadly sympathetic to the initiative and willing to compare sample data with libraries

Some publishers already routinely providing data for local loading,,for others a new concept

Would the ‘default’ post termination access still be via publishers’ sites? Would the archive be permanently ‘open’ or only a ‘dark archive’? What would be the format requirements for content and metadata?

12

Page 13: PECAN Final Report

How would authentication be handled in the archive (ie access only by subscribers)?

Could archive usage data be provided to publishers? There may be a cost for providing the data What would be the ‘start’ date of journal content? Complex to build and maintain Movement of titles amongst publishers will be a challenge

Key issues emerging:– Librarians seemed more interested in the potential value of a registry than

the archive itself– Librarians/academics are more challenged about how quickly they can

get access to content if it is no longer available than who is looking after it/where it is

– Some publishers seemed to suggest that this whole area isn’t a major problem and that existing mechanisms are already in place

– Titles moving between publishers and recording entitlement to unsubscribed titles appear to be the biggest problems faced by libraries and potentially for the registry

– There is still some general uncertainty in this whole area about preservation/archiving and the various services that exist already

– Should the archive be wholly held at a single location or making use of existing archive services

B. From the online survey

25 responses received from 24 institutions (2 from Bath)

Institutions responding:

Kingston Loughborough Bath Sheffield MRC NIMR SurreyRoehamptonBath SpaUlsterWorcesterScottish Crop Research InstituteImperial College LondonNorthamptonAstonBristol Sussex De Montfort Portsmouth Bangor London School of Economics

13

Page 14: PECAN Final Report

Cumbria Birmingham Abertay Dundee Exeter

General responses to questions:

Q. Would you welcome the creation of a Web-based registry recording the access archival rights relating to NESLi2 journals, for each institution participating in NESLi2 agreements? (It is anticipated that the registry would list individual journal titles and individual institutions, showing the periods that each institution is entitled to post termination access for each title)Yes – 25

Q. What do you think would be the benefits of such a registry?

A centralised comprehensive independent registry which can be used as proof of entitlement which would be reliably superior to the institutions and publishers own records, particularly where titles move publisher, change titles or cease.

Q. What do you think would be the challenges / difficulties (if any) in creating and maintaining such a registry?

There would be a tremendous amount of work involved in setting up the registry to agree and confirm title lists and coverage, particularly for deals which have been running for many years and where titles have already moved publisher. It would also be a large task to keep the registry up to date and accurate in future years.

6. Would you welcome the development of a central dark archive of NESLi2 e-journals within the UK HE community? Yes - 25No - 0

There is some concern over whether this would duplicate the work of Portico and LOCKSS and whether publishers would be happy to sign up to another initiative. Although welcomed it is with less enthusiasm than the registry idea.

Q. It is anticipated that for individual titles that have been cancelled, access by institutions would continue to be obtained from individual publishers’ websites. The archive would come into use in situations where events have caused access from publishers’ sites to be unavailable. Would you agree with this approach?

All replies were positive.Although it is felt that the ideal situation would be to have access to all cancelled titles from the same source, as long as it was clear where you could access the cancelled titles from it would be acceptable. Such titles would need to be regularly checked though to ensure that access remained available on the publishers’ sites. There is a concern about the costs involved of having to pay for a further dark archive.

14

Page 15: PECAN Final Report

Q. What do you think would be the challenges / difficulties (if any) in creating and maintaining such an archive?

Publisher cooperation and buy in Competition from other products eg Portico and LOCKSS Further costs for institutions in subscribing to the archive The provision of ongoing funding for the project which would involve a large

investment in time and money both in setting up and running and keeping the data up to date and accurate.

Server space, technology challenges and support

(See Appendix B for additional comments from respondents to each question)

Comparison of subscription data between publishers and librarians:

The aim of the exchange of data was to see whether a small group of publishers and librarians agree on which titles are eligible for post-termination access, as well as the periods they are accessible for. It also allowed us to assess the issues that might cause a disagreement of understanding of rights. Additionally it was intended to allow the project team to consider the potential scale of the Registry and archive. The comparison included assessment of both core ‘subscribed’ journal titles, and also additional titles that are made available in a Big Deal or Collection.

In order to make the comparison manageable for librarians and publishers, we restricted the data to journals subscribed to in 2008 and 2009. The aim was to seek agreement (or not) on rights, if hypothetically a library had cancelled these subscriptions at the end of 2009.

The following are the libraries and publishers for which data was compared:

Elsevier IoPP OUP RSC SAGE Springer W-B

Glasgow √ √

UCL √ √

Cardiff√

Aberdeen: √

√ √

Hertfordshire √ √

Edge Hill: √ √

Kingston√

15

Page 16: PECAN Final Report

Findings from the comparison:

The comparisons were made with only the small group of libraries and publishers, so only give a snapshot of the situation relating to these organisations. However, the results show that in principle the publishers and librarians are broadly in agreement on which content qualifies for post termination access .

A relatively small number of issues came out of the exercise:

In some cases publishers indicated one or two extra titles that were covered by the post termination access rights. These were mainly new titles that the libraries may have missed.

Titles that prove to be databases rather than full text journals do not have post termination access rights, but this is accepted by the librarians.

In some cases titles may be part of an archive that has been purchased through a JISC Collections deal which are available anyway in perpetuity.

Where a society owns a title, it could be the case that the society did not make available a title as part of a publisher’s Big Deal, but then agreed to do so the following year. This means that the years of post termination availability could vary for a title

In addition to titles moving between publishers, another complicating factor will be the extent to which institutions have merged. Whilst we have not seen many examples of large organisations coming together, there have been examples over the years of medical schools becoming part of universities, and this will make the task of establishing post-termination rights more difficult. UCL is a case in point.

It is clear that when scaled up across all institutions and all NESLi2 publishers, the scale of data to be handled (for the registry and in particular for the archive itself) will be very extensive and additionally the complexity of recording varying years of rights, changes of titles etc should not be underestimated.

A range of sample data from publishers and libraries has been provided to Adam Rusbridge and Peter Burnhill, and further sample data can be provided if needed.

Section conclusions:

Our sample group of publishers and libraries are broadly in agreement regarding the journal titles they are entitled to access post-termination and we have no reason to believe that the discrepancies could not be resolved relatively quickly. The most likely reasons for discrepancies appear to be related to titles transferred in or out of a collection or titles whose status in respect of the existing NESLi2 agreement was not fully clear. One institution didn’t realise that they were entitled to post-termination access to unsubscribed journals and simply listed the titles where they had historic print subscriptions.

16

Page 17: PECAN Final Report

Organisations needed some time to compile their data. One institution simply copied the relevant Schedule from the current licence but most appeared to have undertaken some more detailed checking and presented the results on an Excel spreadsheet.

Whilst the number of discrepancies was low and the time taken to gather the data was relatively short, we were only working with a very small sample and a considerably longer period would be required if coverage were to extend to all NESLi2 publishers and participating institutions

All of the publishers in our sample, excluding Elsevier, offer post-termination access to ‘unsubscribed’ journals (in addition to subscribed journals ) covering the period that an institution has licensed this additional content. Five of the publishers in our sample offer the option for institutions to licence different levels of access (Wiley-Blackwell for example offer the Full Collection, an STM Collection and an HSS Collection). These different options and levels of entitlement, when added to the volatility of the data as a result of titles moving between publishers, suggests that the work of gathering and recording the data and keeping it current, will be considerable.

In respect of title transfers, it will be worth following the work of Project TRANSFER which was investigating the feasibility of setting-up a basic alerting service to notify libraries of title transfers as and when they happen

One of the respondents to our online survey, suggested that the subscription agents might still be the most accurate source of information relating to entitlements post-termination and for some publishers this is likely to be the case, especially for data relating to individual core subscriptions.

As we indicated, sample data was only requested for 2009 and 2008. One of the issues to address and decide on in the next phase would be the extent of back years to be included in the registry and archive when they go live

17

Page 18: PECAN Final Report

Section Three

Candidate technical infrastructure for a registry of entitlement and a secure archive

Four major components comprise our proposed information architecture:

1. A post-cancellation resolver service that responds to requests from library patrons and third party serials services, via both a browser service and m2m interoperability, by utilising entitlement and locate registries,

2. An entitlement registry that matches up title information with institutional subscriptions and post-cancellation entitlement,

3. A locate registry to redirect the enquiry towards potential and appropriate sources of supply, and

4. A central archive of content that provides assured continuity of access.

In the following sections, we discuss these components in further detail. Note that this is not intended to be a comprehensive analysis; rather it is intended to highlight the data dependencies, resources and components (both extant and now required). A full

18

4. Central Archive

1. Post-Cancellation Resolver Service

2. Entitlement Registry

E-Journal Registry(supplied by ISSN/

SUNCAT)

Metadata on Subscriptions(supplied by Subscription

Agent)

3. Locate Registry

E-Journal Registry(supplied by ISSN/

SUNCAT)

Metadata on Locations

(supplied by Preservation

Agency)

Service Provider

National OpenURL

Router

LIBRARYUSER

SERIALS SERVICE

Third-Party Resolver

HEFCE / SCONUL

ERM

Page 19: PECAN Final Report

requirements analysis and technical specification would be core activity of pilot implementation.

Our goal through PECAN is to improve availability of accurate subscription information and access post-cancellation. Library systems and mechanisms currently deployed for recording and tracking subscription information are complex and, though they work, may be challenging to maintain. PECAN will not remove all the complexities of data management but we are proposing the entitlement registry as a way to improve the quality of subscription holdings data while reducing the costs and resources required for ongoing library management.

Functional OverviewThe PECAN infrastructure is designed to support four functions:

1. Receive request following failure at / cancellation with original publisher:When a library ceases a current subscription, its library catalogue / ERM is updated to include a link (OpenURL) to the post-cancellation resolver service.

2. Discovery of entitlement:When a request is received, the post-cancellation resolver service identifies the originating institution and the journal volume being requested. The post-cancellation resolver service uses to the entitlement registry to discover entitlement.

3. Locate appropriate source of supply:The post-cancellation resolver service then refers to the locate registry to discover preferred/available sources of supply.

4. Re-direct request for post-cancellation access:The post-cancellation resolver service then forwards the library request to the appropriate source of supply, along with an authorisation token. This model assumes that authentication of the individual has already taken place, most likely via delegation by the post-cancellation resolver service to an appropriate service such as the UK Access Management Federation.

Overview of Components

1. Post-Cancellation Resolver ServiceCurrent library infrastructure is geared to supply access to electronic journal content through the use of OpenURL link resolvers that maintain knowledge of current subscription. In the instances where content is open access, or knowledge of current subscription is held, we expect the current model to satisfy library demand (although acknowledge that these processes are under investigation by the HEFCE/SCONUL Shared Services Initiative).

Current systems appear less well equipped to cope with post-cancellation entitlement where historical subscription information is required, and the model is only now being tested. Introduction of a post-cancellation resolver service is intended to improve the environment, offering a degree of standardisation and greater assurance of stability to libraries and publishers.

19

Page 20: PECAN Final Report

The post-cancellation resolver service must be able to address two central questions: entitlement and location of supply of service. The data dependencies involved are detailed in subsequent sections.

The role of the National OpenURL Router is to provide consistency across the variety of third party resolvers to which different institutions subscribe. The proposed post-cancellation access resolver service could be considered as yet another resolver with which the OpenURL Router could operate, but used primarily in the case of a failed attempt to find an alternative source. One model for implementation of the post-cancellation resolver service will be as an extension of the existing National OpenURL Router.

2. Entitlement RegistryThe entitlement registry would be used by a variety of communities including libraries, publishers and service providers to determine whether access to licensed resources is authorised.

The central record held by the entitlement registry will be a repeatable licensed resource descriptor that comprises the following4:

Title of the licensed resource and its identifier (e.g. ISSN); Name of the publisher responsible for the resource; Name of the agent responsible for this descriptor; Licence conditions for this resource; Service provider access points where the resource may be accessed (included

here to show the relationship, this data element will be stored in the locate registry);

Approved access management technologies for users of this resource; List of licensed institutions, where each instance comprises institution name and

other details, such as site representative details, and registered IP descriptors.

In summary, the entitlement registry will hold information on institutional subscriptions to journal titles, with granularity specified at either volume or year level, and structured in a way so that data can be parsed automatically. Fields must be permanent and repeatable in order to record the history of any change.

When a library enters a contract, the publisher or an agent acting on their behalf will record in the entitlement registry the opening of that subscription as part of that contract process. The registry will treat ‘open’ records as continuing subscriptions. A publisher or agent acting on their behalf will need to explicitly record the termination of a subscription in the entitlement registry, again as a process of contract. In order to record the duration of contract, the following fields are required in the license descriptor.

Opening date of contract; Opening volume and/or year of entitlement; Closing date of contract; and Closing volume and/or year of entitlement (if any).

4 Note that the fields listed are adapted from the unpublished report: JISC Committee for Authentication and Security (JCAS) — Scoping study for interim DNER authentication developments; Sandy Shaw; Computing Services; The University of Edinburgh; September 2000

20

Page 21: PECAN Final Report

Metadata on subscription action will be supplied from the publishers, or an agent acting on their behalf. The ISSN may be able to supply metadata on title information. PECAN may be able to contribute to the workflow for updating the publisher information as registered with the ISSN-IC, however it may be that existing initiatives such as Project TRANSFER can more usefully play a more central role with this. Reconciling data from different sources will be a challenge and needs further analysis, but is not expected to be insurmountable. For example, it may be possible to build on the activity of the International Committee on Electronic Data Interchange for Serials, ICEDIS (http://www.icedis.org/), which has set standards for the electronic exchange of data between agents and publishers. A pilot implementation would require liaison with publishers and acquisition of trial data in order to see what conversion might be needed.

3. Locate RegistryOnce post-cancellation access entitlement has been established, it is necessary to discover an appropriate source of supply. In current library infrastructure this task is achieved through the use of link resolvers; commercial vendor systems that utilise a proprietary “knowledge base” of links to scholarly content with metadata specified at an appropriate level of granularity (i.e. from title down to article). The accuracy, comprehensiveness and currency of the knowledge base data are critical for library take-up.

A locate registry is proposed within PECAN to resolve requests for access towards potential and appropriate sources of supply. The scope of the locate registry is limited: any service provided by PECAN is not intended as replacement for current library infrastructure such as link resolver systems. However, within this limited scope the contents of the locate registry can be considered as somewhat analogous to link resolver knowledge bases (KBs). The support of publishers as providers of location information will be critical here. Commercial link resolver systems are valued for the timeliness and accuracy of their information, and indeed much resource is spent on ensuring this quality.

Further analysis is required to consider the extent to which existing practices can be reused for a locate registry. Agreement to provide location data is needed from publishers, subscription agents and third party initiatives, and prior investigative projects have indicated publisher reluctance to do so5. Libraries may be hesitant to commit to a new service until it has an established user base, and so service level quality assurances are needed. Indeed, the quality of service must enhance current technologies and efficiencies and at minimum must keep abreast with current developments in the field. This applies not only to the data quality of the locate registry contents, but also the functionality provided by a post-cancellation resolver service.

The extent to which location information is readily available from alternative sources of supply, such as dark archiving initiatives, is not yet known. If this data is currently unavailable, this may be because there has not yet been high demand for location information. Link resolver systems to date have focused on current subscription, though

5 A low-cost OpenURL resolver for UK HE/FE: a scoping study; Leah Halliday, Andrew Bevan, Tim Stickland and Christine Rees; EDINA, May 2008

21

Page 22: PECAN Final Report

this may change as the quantity of orphaned titles increases. Dark archiving initiatives are new services, and much of the focus of their activity to date has been on encouraging publisher participation and collecting content. Introduction of a locate registry dedicated to post-cancellation access may stimulate best practises and lead to wider accessibility of location information.

A PECAN entitlement and locate registry may contribute to improving the quality of OpenURL resolver knowledge bases. A recent scoping study proposing a low-cost OpenURL Resolver6 surveyed the library community to understand desired functionality associated with such a resolver, and the results “reflected the frustration within UK HEI libraries that they are unable to obtain accurate and current information about the e-journals packages negotiated for UK HE as part of the NESLi2 initiative.”

One method of implementation may be through a central UK-focused OpenURL link resolver. In the interests of data reuse, activity during pilot implementation would include an analysis of how current link resolver vendors may utilise an entitlement registry, and whether a locate registry could be used to improve their service quality.

Transfer of titles between publishers can result in a change in the location of access, and this must be reflected in library systems. The process for provision of post-cancellation access following transfer between publishers has improved following introduction of the Code of Practise established by Project TRANSFER. It will be necessary for the PECAN locate registry to be updated to reflect new and replacement sources for post-cancellation access as they become available. It may be possible to integrate this update procedure as a component of the Project TRANSFER workflow; however it may be that PECAN is not the responsible body. For example, PECAN may gather information on titles and publishers from agencies such as the ISSN-IC. In this event, it would be desirable that the original ISSN records were updated directly.

4. Central ArchiveThe extract below from Clause 8.5 in the NESLi2 Model License outlines conditions for provision of post-cancellation access to journal content. One mode of access to back content is via a central archiving facility operated on behalf of the UK HE/FE community, following supply of archival copies.

After termination of this Agreement […] the Publisher will provide […] the Licensee […] with access to and use of the full text of the Licensed Material which was published and paid for within the Subscription Period,

By

(i) continuing online access on the Publisher's interface, (ii) a central archiving facility operated on behalf of the UK HE/FE community, or or other archival facility (iii) supplying the Licensee with archival copies

6 ibid.

22

Page 23: PECAN Final Report

A central archive of electronic journal content for use by UK Higher Education institutions was discussed at the PECAN consultation event in October 2009. There was shared priority across the publishers and librarians present, that researchers and students should have as much ease of access to back copy post-cancellation as to currently subscribed content. The challenge was to devise and provide a location facility that addressed the issue of continuity of access: resolution to assured source of supply.

The necessity of a UK central archive alongside currently available sources of supply needs further debate and consideration. Exactly how much responsibility publishers are willing to take for provision of post-cancellation access needs to be clarified. International initiatives have emerged to safeguard long-term access to content, however, currently these tend only to be used in the event where orphaned content is no longer available from any other source.

Policy makers in the UK need to assess whether current models of supply (via publisher and/or dark archive initiatives) are sufficient, or whether further assurance of access is needed by locating content within UK-based systems and infrastructure, and operated under local policy control.

Participation by UK Higher Education institutions in available preservation services is not yet comprehensive. Knowledge of who is doing what is addressed in PEPRS. Among UK HE institutions there has been interest in the introduction of a central UK archive. A central archive would ensure the UK takes appropriate responsibility for providing long-term access to content. Indeed, JISC and its data centres have much history in maintaining centralised data management initiatives, and central funding may prove to be the most cost-effective and stable method to support library access to archived material.

Existing preservation initiatives demonstrate that electronic journal archiving is challenging. Building on existing infrastructures may reduce the complexities associated with establishing an archive. One approach may be to exploit the LOCKSS technology, software to ensure the long-term accessibility to electronic journal content while respecting business elements such as subscription rights, even if the original publisher ceases to exist, a subscription is terminated, or the already acquired content becomes damaged. An attractive aspect of this software is that it is designed to support different organizational models for use, allowing groups and networks of institutions acting together to take responsibility for preservation of content.

Building upon existing software may also permit reuse and extension of existing publisher workflow processes. For example, seven of seventeen NESLi2 publishers are already participating in CLOCKSS, a dark archive of electronic scholarly content jointly governed by publishers and librarians on behalf of the global community. If a central archive were to be established in the UK, it may be attractive for publishers to use technologies and workflow processes already in place.

If a central archive were deemed necessary, the central archive should in the first instance comprise material from publishers participating in NESLi2. The involvement and support of JISC Collections can be expected to encourage publisher participation; however the terms under which content shall be accessed will need to be explicitly defined in collaboration with publishers, as a component of governance.

23

Page 24: PECAN Final Report

Organisational and governance models for any archive would need more thorough investigation than can be achieved within the scope of this report. We advocate JISC Collections having a central lead role in the management of the archive. It may prove more feasible for the development and maintenance activities to be delegated to external organisations, for example the JISC National Academic Data Centres at EDINA and Mimas.

24

Page 25: PECAN Final Report

Conclusions and Recommendations

Conclusions

This scoping study conducted as the PECAN project has confirmed that post-cancellation access has become a very important issue for both libraries and publishers, providing a framework for policy and action. In a climate of severe downward pressure on library budgets this will become even more significant, especially with the retreat from the Big Deal model.

The work undertaken in the PECAN project indicates that there are examples of methods to provide post-cancellation access, but there are significant differences in the perception of information held by publishers about individual institutions journal entitlements to be addressed, at machine-to-machine level to support ease of end-user access as well as at the level of the librarian. This information is also expected to be subject to change over time.

In an environment dominated by the need for higher education to make large scale cancellations to their current subscriptions, the attitudes of publishers towards the provision of access to cancelled titles via their web sites may change.

Policies in this area should be monitored over time for evidence of negative impact on the ability of universities and colleges to secure post-cancellation access from the publisher interface for their patrons.

There was considerable interest in the survey and focus group of publishers and libraries, especially from the library community, in the provision of systematic, accurate and up-to-date information on subscription status and entitlement. Termed a registry of entitlement, a subscription database with a record of history, was seen as important for:

1. The clarification of archival rights for the titles in journal agreements, post-cancelation, especially to ‘unsubscribed’ titles that were bundled into Big Deals.

2. The provision of an accurate record of archival rights to society or transfer journal titles, especially those made available by NESLi2 publishers but which fell outside a NESLi2 agreement.

The provision of accurate information regarding for the location (URL) of online access services for such content, including archived journals.

The provision of a central archiving facility was also mooted but concern was expressed about the possible costs associated with such a facility and the burden that this might place on both the JISC’s budget and their own institutional budgets should they be asked to contribute. Publishers also appear to have concerns about a central archiving facility and their willingness to deposit their journals within it unless UK HEIs moved to a national purchasing model where there was universal access to their journal holdings.

It is not clear that the demand yet exists for the establishment of a central secure archive for NESLi2, but nor is it clear that current initiatives and services sufficiently address the needs of the entire UK HE community. Further investigation is recommended into the criteria that a long-term post-cancellation service should satisfy, in order to assess and benchmark the suitability of current offerings.

25

Page 26: PECAN Final Report

Recommendations

The following recommendations are made about the Registry of Entitlement and a Locate Facility, for which there is both requirement and support:

1. JISC should undertake and fund urgent investigation towards the establishment of such a Registry of Entitlement, beginning with a real-world test Registry based on existing NESLi2 Agreements;

2. Equivalent investigation and test implementation should be made into supporting mechanism to locate and assist access to online content by end users JISC should give this priority within the context of its strategy for digital library

provision, including the use of registries and OpenURL resolvers.

The following recommendations are made about the central archive for NESLi2 journals:

3. JISC should maintain a watching brief over developments in publisher policies towards access fees for post-cancellation access;

4. JISC should continue to review the participation of both libraries and publishers in third party archiving services;

5. JISC should assess as a matter of urgency whether current models of post-cancellation access (via third-party initiatives) are sufficient, or whether further assurance of access is needed by locating content within UK-based systems and infrastructure, and operated under the policy control of UK HEIs .

The following recommendations are made with respect to the provision of post-cancellation in NESLi2:

6. JISC Collections, building upon the NESLi2 model licence, should seek greater clarity from publishers in negotiations over the status of the archival rights applying this to all journal titles included within a given journal agreement;

7. Steps should be taken to ensure that all publishers have stated policy and provision for the deposit of their journal collections with a trusted third party archiving organisation, with clear and satisfactory arrangement for long-term access The JISC-funded PEPRS project is addressing how this information might be

confirmed by the archiving agencies. This is an important first stage in the guaranteeing access to journal materials;

8. Building upon the existing clauses in the NESLi2 model licence, similar steps should be taken with respect to archiving arrangements for content, and for access arrangements, following cancelation of subscriptions All Big Deal agreement should include provision for post-cancellation access

to ‘unsubscribed’ titles as well as ‘subscribed’ titles;9. JISC Collections should assess the needs of the UK HE community in order to

specify what is required of archiving organisations, and to achieve consensus on criteria for one or more designated third party archiving organisations as default arrangement in its licensing work JISC Collections should endeavour to put in place a national agreement with

the third party selected. Such a national agreement would ensure that all UK HEIs were covered by the third party’s archiving services;

10. JISC Collections should continue to explore licensing agreements that offer all UK HEIs access to journal agreements, in line with the collaborative purchasing model used by the Scottish Higher Education Digital Library (SHEDL).This

26

Page 27: PECAN Final Report

activity would create the licensing conditions that would support the establishment of a central dark archive.

27

Page 28: PECAN Final Report

APPENDIX A

Pilot for Ensuring Continuity of Access via NESLi2 (PECAN):

An Exploration of Issues

Status: Draft, v.03

June 2009

1. Introduction

This sets out a proposal for a short pilot project (entitled PECAN) to investigate what is needed to ensure continuity of access to content licensed under NESLi2 in the event of termination of the subscription agreement (Clause 8.4, detailed below). There is some urgency to this as the prospect of journal subscription cancellations beyond July 2009 is real, and may be large-scale beyond July 2010.

Clause 8.4 of the NESLi2 model licence specifies the context in which continuity of access is permitted. 8.4 After termination of this Agreement (save for a material breach by the Licensee

of its obligations under this Agreement) the Publisher will provide (at the option of the Licensee) the Licensee and its Authorised and Walk-in Users with access to and use of the full text of the Licensed Material which was published and paid for within the Subscription Period, either by i) continuing online access to archival copies of the same Licensed Material on the Publisher's server which shall be without charge; or ii) by supplying archival copies of the same Licensed Material in an electronic medium mutually agreed between the parties which will be delivered to the Licensee or to a central archiving facility operated on behalf of the UK HE/FE community or other archival facility (excluding an archival facility of a STM publisher) without charge; or iii) supplying without charge archival copies via ftp protocol of the same Licensed Material.

The purpose here is to describe in outline a six-month to one-year project which will investigate, design and test the technical infrastructure and associated policy and procedure for services that enable Clause 8.4 to be put into effect. At least two components are envisaged: a registry of entitlement (maintaining the journal content that has been subscribed by libraries via NESLi2); and a secure archive (providing access to back journal content).

28

Page 29: PECAN Final Report

1.1 Registry of Entitlement

Were such a registry to be feasible then it might assist publishers with respect to their contractual responsibility and reassure libraries that their entitlement to content of journal titles is maintained despite potential changes in service providers. With authorisation managed by a central registry, continuity of access could be provided through one or more appropriate means of service delivery. The characteristics and feasibility of a central entitlement facility, operated on behalf of the UK HE/FE community, are to be investigated in PECAN.

Access to subscribed content requires authentication and authorisation, which in turn requires access to one of more knowledgebase containing information on current subscription status. Although there have been proposals in the past to establish a global licensing table, the management of subscription information remains a challenge. A number of commercial companies, who provide OpenURL resolvers (for the ‘appropriate copy’, perhaps the closest analogy to this problem), maintain a knowledgebase but the focus is on current status; post-cancellation access requires history.

Not all subscription arrangements grant archival access despite the clause of the NESLi2 license, or may limit these: for example, subscription deals may include short-term access to only ten-years of back content. A library has interest in access to content for a given journal title and there needs to be clarity about access continuity rights and mechanisms when titles change hands between publishers or two publishers merge.

These concerns might be mitigated by a central registry that would store and provide information on authorisation by virtue of entitlement through a NESLi2 license. This registry would not preclude other steps being taken to improve library access to information, such as a push towards back content older than X years being made available under an open access licence.

1.2 Secure Archive of Journal Back Content

The NESLi2 licence specifies three means for enabling continuing online access to back content (archival copies): (1) via the Publisher's server, without charge; (2) via a central archiving facility operated on behalf of the UK HE/FE community, following supply of archival copies; (3) by supply of archival copies to each Licensee. The characteristics and feasibility of a central archiving facility, operated on behalf of the UK HE/FE community are to be investigated in PECAN.

Funded by JISC, publishers would have the opportunity to deposit material in a secure and access controlled environment without incurring additional annual costs, in order that patrons of previous subscribing institutions could access this material. Authentication could occur via the UK Access Management Federation with the PECAN Registry of Entitlement used for authorisation.

It is also envisaged that this central facility might be both real and virtual, meaning that a publisher could opt to make separate provision for assured access to back content, directly or with a third party (typically a trusted archiving agency), having met certain pre-agreed terms and conditions for longevity. This could be managed as part of the entitlement registry but this relates more directly to matters of digital preservation and delivery of access.

29

Page 30: PECAN Final Report

2. Scoping a Pilot for Ensuring Continuity of Access via NESLi2 Content (PECAN)

We envisage that PECAN would be a service composed of two parts. The PECAN Archive and the PECAN Access Management Service.

PECAN is designed to ensure that NESLi2 access continuity entitlement is well managed in a stable UK-led trusted and secure environment. This can be achieved through a central registry of institutional subscription rights to content, specifying the titles and date ranges for which access continuity rights are granted.

NESLi2 publishers already maintain institutional subscription information but access

to these records are limited. There are concerns that a publisher may not be able to assure access to these subscription records following mergers or closure.

We propose investigating the role that could be played by a central OpenURL

resolver, limited in scope to NESLi2 publishers, for the maintenance of post-cancellation entitlement information.

We propose investigating the use of ONIX for Subscriptions to assist with aggregation of subscription information.

We propose JISC Collections as the organisation responsible for maintaining access continuity entitlement information, acting on behalf of UK libraries, and cooperating with NESLi2 publishers to do so. The daily activities associated with this may be outsourced to an appropriate UK service provider, but we promote the core responsibility lying with JISC Collections.

The PECAN Archive will be a UK-led standalone archive for the preservation of scholarly content. It may be possible for the archive to build upon existing solutions such as the LOCKSS infrastructure. Sufficient development expertise for the PECAN Archive should be in place to contribute to the ongoing stability of the software platform.

In the first instance, we expect the PECAN archive to cover only those publishers participating in NESLi2 licences.

The distinction between the PECAN Archive and the PECAN Access Management Service will allow publishers and libraries to designate a preferred agency through which access continuity provision occurs. We do not believe that the PECAN Archive will or should be the only provider of post-cancellation access, but availability over the long term needs to be assured.

PEPRS, the JISC-funded project activity being carried out by EDINA and the ISSN International Centre, currently has focus on digital preservation of e-journal content, not continuing access to back content.

3. Technical and Other Considerations

-- Individual Access via Authentication/Authorisation --

30

Page 31: PECAN Final Report

It is envisaged that the PECAN Archive would be regarded as a service offered by a service provider (SP) registered with the UK Access Management Federation. For the purposes of the PECAN project, EDINA would be that SP. Individual (authentication) access would occur via their institution, as a registered IdP, with authorisation via the PECAN entitlement registry.

-- Subscription Information --

The key requirement for the registry is to host information on institutional subscription entitlements, which might best be handled in the medium term through subscription holdings information received directly from the content supplier, possibly using the ONIX for Subscriptions XML format.

For the immediate future, there are challenges about information availability. Subscription information is not readily available from publishers. Commercial companies providing OpenURL information to libraries (perhaps the closest analogy to this problem) have information, and may be persuaded to assist. The major subscription agents may also be persuaded to assist.

It would be challenging for us to aggregate this hard-to-discover information on a low budget, and we may be viewed as competing with these commercial ventures. It’s not clear how many publishers are ONIX compliant.

Publishers seem unwilling to provide data, but it’s not clear if they don’t want to provide the information or that technically they cannot. It may be that publishers won't provide the data because they don't currently have mechanisms to deliver that data in place, and to do so would require some investment on their part. With the support of JISC Collections, it may be within scope to require delivery of subscription information as part of the NESLi2 license. This could certainly be assessed for feasibility.

31

Page 32: PECAN Final Report

APPENDIX B:

List of all additional comments from libraries responding to online survey:

Would you welcome the creation of a Web-based registry recording the access archival rights relating to NESLi2 journals, for each institution participating in NESLi2 agreements? (It is anticipated that the registry would list individual journal titles and individual institutions, showing the periods that each institution is entitled to post termination access for each title)

Useful for staff outside serials and other back office depts as well as those within. It would make it easier to ascertain what we should have access to and help to

make sure we still had access to the content. I think it’s an excellent idea. Brilliant As long as it doesn’t overlap with other proposed projects, eg PEPRS Vast improvement on the current situation Providing accurate data can be supplied by publishers/agents

Q. What do you think would be the benefits of such a registry?

Kingston – I think it is absolutely essential and would allow material which can be quite hard to find to be retrieved quickly and easily.

Loughborough – It would assist with records keeping, especially in terms of selecting print volumes for discarding

Bath – The potential to easily access a comprehensive list of material to which we have archive entitlement that could then be checked against our in house system.

Sheffield - One stop shopping MRC NIMR – It would be an independent register of rights. In the event of

mergers/acquisitions it would be evidence in case of any disputes or loss of records with publisher.

Bath – To save us having to decode the information individually – publisher commitments vary widely but the NESLi2 packages ought to mean that at least for any given publisher the rights should be identifiable.

Surrey - This would cut out a lot of legwork in looking through individual licences when we have to make a decision that involves assessing archival rights (eg cancellation of a title).

Roehampton – Easier to find back issues. Wouldn’t be so insecure about access and storing print copies to ensure this.

Bath Spa – It would be a clear method of maintaining such information into the future.

Ulster – In relation to titles we currently have access to or think we should, it is becoming increasingly difficult to know whether we do and why as journals change hands (and titles) and as contacts change. After a cancellation it would be very difficult to know what we ought to have. The registry would share skills and knowledge needed to reach agreement.

Worcester – Clarity of what we are entitled to have access to.

32

Page 33: PECAN Final Report

SCRI – Knowing which titles were included in which deals and the years applicable is difficult

Imperial – At Imperial we find the job of collecting data on our post-termination entitlements fiddly and time consuming and I’ve always had nagging doubts over how far we can rely on the information provided by publishers. To have this done at national level should improve the reliability of the data and will save us staff time.

Northampton –Matches current trend towards federated searching. Covers complexity of the market, where titles move between vendors and potential problems with publisher’s sites.

Aston – It is very hard to keep track of this as individual institutions so it would be labour saving. Also to have an overview would facilitate accurate delivery either by publisher or someone else. Also should cover publisher changes.

Bristol – It would provide a third party confirmation of archival rights. Sussex – It would be useful to have all the information located in one place. De Montfort – Clarify what titles we have rights to. Clarify time periods. Enable

accurate catalogue and link resolver information. Portsmouth – To ensure retention of access rights over time and irrespective of

personnel changes in individual institutions Bangor – A centralised location for this information would save lots of

investigative work that we currently have to undertake. It would enable us to make decisions regarding our subscriptions on a more informed basis.

LSE – Transparency, understanding Cumbria – It would make things easier to us when trying to find out what we had

access to historically Birmingham – At the moment, even within NESLi2 deals which should be fairly

standardized, there is a lot of history relating to an institution’s particular entitlements that might differ from the norm and requires some checking with the publishers to ensure that our entitlements are correct. Often the publishers struggle to find the information themselves. So the registry part would be a huge improvement.

Abertay Dundee – Individual institutions won’t have to provide their own solutions. Also won’t need to liaise individually with suppliers re hosting data, authentication etc.

Exeter – Useful to confirm cancellations. Help agents/library set up ERM information accurately.

Q. What do you think would be the challenges / difficulties (if any) in creating and maintaining such a registry?

Kingston – Are publisher policies clear? Do they know what they offer now and what they used to offer? Are they happy to make the information available in this way?

Bath – Accuracy of data is a constant challenge, for instance handling publisher changes.

Sheffield – Currency, no good without. Getting publishers to be clear MRC NIMR – it would be a nightmare to set up, to agree all rights and dates. Bath – Title changes/transferred titles. Different package options in different

years

33

Page 34: PECAN Final Report

Surrey – Keeping on top of changes due to title transfers, publishers licences terms over time

Roehampton – keeping it accurate and up to date Bath Spa – People comment that publishers don’t know what titles they should

be providing as part of NESLi2 deals and librarians have great difficulty getting the correct data.

Ulster – Tracking licences and rights from the earliest days of NESLi – what about deals where we have pre-NESLi rights? Title and publisher changes.

Worcester – time taken to set it up and to make sure it is kept up to date and accurate. The ability to obtain correct information from publishers especially when titles move between them.

SCRI – Titles that change their names and move from one publisher to another with changing access rules.

Imperial – setting it up could involve a fair bit of work. Northampton - Challenge of including all entitlements: publishers and institutions

often have difficulties agreeing these when a number of units within an institution may be purchasing content. Challenge of keeping up with the cancellations, new subscriptions, changes in licences over time.

Aston – Collection the information as a one off and maintaining it accurately. Libraries and publishers often take several iterations to agree the situation. Also journals change publisher and deals can have different content for different institutions if you add title(s).

Bristol – Establishing what each individual institution has rights to as some NESLi2 deals don’t specify the content each institution has access to.

Sussex – Accuracy, keeping it up to date, changing formats etc De Montfort – Moving walls. Publishers unwilling to maintain access to what we

have paid for. Portsmouth – Assembling accurate information in the first place; maintaining this

over time Bangor – Probably obtaining accurate information for each institution in the first

place. This is where we struggle at the moment, often having to contact our agents, publishers etc to try to find the data we need.

LSE – Finding an automated way of establishing titles, years of coverage. Cumbria – keeping it current and relevant Birmingham – It would need to keep on top of title transfers and journal

cessations and title changes which would be difficult. Also although the licence states that perpetual access may be available to subscribed content from a set date (eg 1997) it may be that because some titles may have moved to the publisher post 1997, we can’t use 1997 as a block date to inform activities like print issue disposal as it could mean disposing of content that we don’t actually have post cancellation access to. So the system would need to record this information.

Abertay Dundee – Maintaining accurate data about access dates. Change of policies and options (despite agreements being signed) when journals move between publishers.

Exeter – Getting all publishers to participate/release data. Time if libraries required to provide data/check accuracy.

Q. Would you welcome the development of a central dark archive of NESLi2 e-journals within the UK HE community?

34

Page 35: PECAN Final Report

Sheffield – need to protect our investment in NESLi over the years MRC NIMR – This is something that is hard for small institutions to justify, but

there is a need for it. Surrey – I would be very interested to know how this overlaps with Portico – as

we’re just about to sign up for that. Bath Spa – as long as it was working with other similar initiatives Ulster – less sure about this than the registry SCRI – with the best will in the world publishers may find it difficult to maintain

the archive Bristol – While we subscribe to LOCKSS and Portico an alternative covering core

content would be welcomed. LSE – Again as long as there is no duplication of effort for us, eg we currently

maintain a LOCKSS box and pay for Portico Birmingham – Current systems that archive this content are on a paid basis (eg

LOCKSS and Portico) and it is difficult to assess comprehensive these services are.

Exeter – important if publisher fails

Q. It is anticipated that for individual titles that have been cancelled, access by institutions would continue to be obtained from individual publishers’ websites. The archive would come into use in situations where events have caused access from publishers’ sites to be unavailable. Would you agree with this approach?

Loughborough – That would be OK, although it might be good to be able to obtain the different publishers titles from the same source

Sheffield – as long as archive info makes that clear Bath – As long as the registry of entitlements was in place to back up claims to

perpetual access Surrey – It’s easier for users if frontfile and backfile are in the same place, unless

there is good reason for them not to be. Roehampton – If we have to pay a platform or access fee to the publisher and

then have to pay for the dark archive as well this could be expensive Bath Spa – Fine if it works. I’ve had problems in maintaining access to cancelled

titles in the past. Worcester – That would be a sensible idea provided that the material was kept

available on the publishers websites; it is where users are used to going to to obtain content.

SCRI – I agree the practicalities make this the best option but having all cancelled titles in one place might be easier for librarians.

Aston – It adds long term security and supports the e only strategy De Montfort – Initially although ideally I would like to see a central repository

holding all material paid for, but not under the control of the publisher Birmingham – It is a last option. I assume that this archive would in any case

only contain content where publishers have already agreed to the perpetual access clause and would provide access on their website. We find the registry option more of a benefit at this point.

Abertay Dundee – It would be more useful to have these individual titles included. If they are included in the dark archive automatically as implied then that should be possible somehow.

35

Page 36: PECAN Final Report

Q. What do you think would be the challenges / difficulties (if any) in creating and maintaining such an archive?

Kingston – Would a cost be involved? Other similar schemes such as Portico ask for reasonably high fees for providing this kind of archive access to a small number of titles. Maintaining it would depend on issues such as knowing in time what access has been lost and who is entitled to it.

Loughborough – Persuading the publishers Sheffield – Competition from other products, eg Portico MRC NIMR – Persuading the publishers to contribute content to yet another such

initiative, when Portico, LOCKSS, British Library etc etc already exist. Bath – Relying on publishers to provide trigger events. Making institutions aware

of content becoming available. Surrey – Size required for so much data, data integrity over time, migration to

new formats over time and access management to a very granular level. And gaining the agreement of publishers especially those who have already signed up to CLOCKSS and Portico and might think that was enough.

Roehampton – Cooperation from publisher. Providing access quickly when required. Would need clear terms and conditions about when it could be used.

Bath Spa – Getting publishers agreement – not overlapping too much with other initiatives in this area.

Worcester – The time taken to obtain the data and keep the information up to date and accurate.

SCRI – I can see publishers and PECAN having a few differences of opinion over what should be provided and what should not.

Northampton – Changes in titles published by an individual publisher and in content of deals over time.

Aston – Publisher buy in and economics Bristol – Ongoing funding as this would have to be a long term strategy rather

than a time limited project. De Montfort – Server space, admin costs, publishers. Portsmouth – Its complexity, given the inevitability of changes over time; but this

is the norm for e-resource management LSE – Getting publishers to agree to hand over content Birmingham – Liaising with publishers on post termination access entitlements.

Longevity and preservation of access/download formats. Abertay Dundee – Technology challenges; publishers changing the rules, tech

support for loss of access/servers down etc Exeter – Keeping accurate records, up to date costs.

Any other general comments you may wish to provide:

Kingston – A welcome study into two great proposals Loughborough – There would appear to be overlap with Portico and LOCKSS but

I still think this would be worth pursuing as I believe that the UK should follow its previous history of providing centralised preservation strategies. Hopefully this

36

Page 37: PECAN Final Report

could eventually become part of a national dark archive perhaps in conjunction with the BL.

Sheffield – Is this affordable? Bath – This sounds very worthwhile – building on the concept of CLOCKSS, but

for this specific and definable set of agreements. The registry of entitlements is probably of higher value, however, as this is generally more problematic than total loss of content.

Roehampton – Are links envisaged with Portico and LOCKSS? Ulster – Is anyone talking to subscription agents? They can sometimes help us

out in sorting out what we have had access to in the past.

37