DMU Library & Learning Services moved to a new form of access management with regards its online resources and electronic journals in the summer of 2011. This service is commonly known as Single Sign On and replaced the former authentication model of issuing individual Athens usernames and passwords to DMU Library patrons to allow them off campus coverage to subscribed online content. DMU Library users can now access the majority of the library’s subscribed eResources and eJournals offsite by using an institutional login; this institutional login provides entry to most other DMU online systems and services (e.g. Blackboard and university email accounts). The transfer to Single Sign On removes the need for DMU students and staff to remember a separate (and extra) login credential when utilising library online resources remotely.
The task of managing and implementing this authentication transition was the responsibility of the library’s Content Management, Planning & Innovation (COMPI) team. The transfer work was multi-faceted, involving many technical changes to database and journal access points across different systems and platforms (e.g. library website, Find it @DMU links). The library liaised with a high number of publishers and data providers in the run up to the summer of 2011 in an attempt to address the access transfer in as seamless a manner as possible (this work continues as I blog!). In many cases the switchover work was done in good time and without too much hassle, but there were a minority of publishers who took longer to make the necessary admin and/or technical alignments to allow the Single Sign On process to operate effectively.
I thought I’d take the time to briefly blog about one issue which has caused problems with the implementation of an authentication method such as DMU’s Single Sign On process. The concern is publisher subscription numbers. I think this has been an on-going problem for many academic libraries over the years – it seems to be akin to publishers in the past having various delivery addresses on file for a university when most journals were purchased in print form (“that site library closed years ago; what do you mean you are still sending issues there?”). You would think the proliferation of online journals over the past decade would have made the ‘delivery’ of electronic content that much easier, but apparently not. Most publishers assign an institution with a subscription or reference number when said institution makes a journal or database purchase. In the case of journal subscriptions, it is not unusual for an academic library to have tens of journal orders with one publisher. In an ideal world, you would think that all of an institution’s journal subscriptions would be positioned under a single subscription number, and that an institution’s admin details and authentication configuration would match this publisher subscription reference (meaning all content purchased is associated with the correct customer). Unfortunately, this appears not to always be the case. Subscriptions administered under different subscriber numbers may not all be aligned to be covered by the library’s access set up with the publisher, meaning some purchased content potentially hangs out in some sort of virtual ‘limbo’ which library users and staff cannot access. I have even known for a publisher to match the library’s IP address (allowing on campus coverage) for some subscription reference numbers, but not have the corresponding remote access criteria to match. Therefore the journal is accessible via a library or university PC, but not available to students/staff viewing from home.
The knock-on effect of multiple publisher subscription numbers is that a library can believe it has set up an authentication/access mechanism correctly with a publisher, but still receive end-user enquiries stating that they cannot access a particular journal or journals. I think further work is needed by some publishers to address this type of admin anomaly. If the academic library has provided the correct details with regards its access management set up (whether it be via Athens, UK Access Federation, Shibboleth or purely IP) to the content provider, then surely it is up to the provider to ensure the library has full access to the content it has purchased?
I would be interested to hear if other libraries/institutions have had experiences of similar problems with publisher/agent references to those described above. They don’t have to be negative experiences either…!