Framework for Improving Link Resolver Systems
When a user conducts a search, either in a traditional A&I database, a federated search, Google Scholar, or clicks through a reference link via the CrossRef system, the essential metadata about the object is passed via an OpenURL to a link resolver system, such as those available from ExLibris, Serials Solutions, Innovative Interfaces, and Openly Informatics. At the heart of each of these systems is a core “Knowledge Base” which provides the context to the OpenURL, comparing it to issue availability data, library holdings information and providing a variety of linking options to the content that was searched. The user is than directed to the content that is available to the content available through his/her institutional subscription.
While these systems work extremely well in the vast majority of cases, they are not without significant inefficiencies and inaccuracies. Much of this is due to the complexity of the distributed supply chain of this information. Link resolver providers (each of which has its own data system, structure, and ingest methodology) receive information from publishers and subscription agents, who provide data on publication release, collections, locations, etc. Frequently, this transfer process requires normalization and quality control review, which add to the complexity and opportunity for error. The library in turn needs to provide holdings and subscription information from their own library systems in order to customize the resolver to match their holdings.
Through a series of conversations with publishers, resolver-systems suppliers and librarians, the research has pointed to some issues and barriers that are inhibiting the deployment and use of these systems. Among the issues identified by SIS were: a lack of awareness that significant issues persist and a lack on cooperation in solving those issues; inaccurate or incomplete data; a lack of procedures for transferring titles; lack of data format and transfer standards; and a communal responsibility for data quality. While OpenURL compliance is growing rapidly, there will need to be broader understanding of the role of OpenURL and how it interacts with other necessary information transfers to facilitate the discovery and delivery of content.
Initial recommendations were suggested and may be explored by UKSG and the community. Much like Project COUNTER, a code of practice might be developed which will address Knowledge Base compliance regarding which information is provided and it what formats. Such a code of practice might certify compliance in areas of format, delivery method, timing and OpenURL compliance among the key organizations in this process, publishers, subscription agents, and resolver suppliers. There might also be areas of standards, which could be developed or expanded to improve this information exchange, such as current work led by EDItEUR on ONIX for Serials Holdings, or possibly a NISO SUSHI equivalent for holdings information.
The final report will be provided to the UKSG Board during their May meeting and the report will likely be posted to the UKSG website sometime shortly thereafter. A summary article is also being prepared for the July issue of Serials. Hopefully, as many other similar UKSG research projects have done, this work will lead to significant outcomes that will improve information exchange.