NASIG Newsletter

22nd CONFERENCE
VISION SESSION

A New Approach to Service Discovery and Resource Delivery
Daniel Chudnov, Library of Congress, Office of Strategic Initiatives
Reported by Yumin Jiang

Daniel Chudnov, formerly of the Yale Center for Medical Informatics, and now of the Office of Strategic Initiatives at the Library of Congress, gave an eye-opening presentation on using COinS and unAPI to facilitate finding and citing information resources, and to integrate scholarly library resources with innovative Web resources and applications. 

The focus of Mr. Chudnov’s presentation was simplicity.  Using the digital media player iTunes as an example, Chudnov asked why libraries cannot work like iTunes, which permits its users to easily connect with each other and share music.  Even with OpenURL and link resolvers, he explained, it still takes many clicks for a patient and savvy user to get from a journal citation to the actual full text.  In addition, there is an apparent disconnect between library resources and many of the Web 2.0 websites and applications.  OpenURL is difficult to understand, inconsistently implemented by information providers, and requires service pre-coordination.  How can OpenURLs be improved to help users find and use library resources?  How can library catalogs/websites and other great Web resources and applications be connected?  Chudnov thinks that new standards such as COinS and unAPI will be able to address these issues.

COinS, acronym for ContextObject in Span, is a specification to render OpenURL to HTML.  This allows client software to retrieve bibliographic metadata and to use an OpenURL resolver to find a mediated link.  The principal advantage of using COinS, rather than giving a static OpenURL, is that the client can determine which resolver to use.  For example, a Yale scholar visiting another institution will be able to access Yale-subscribed resources via Yale’s link resolver instead of the host institution’s.  OCLC has recently established the OpenURL Resolvers Registry.  It includes an OpenURL resolver registry for user input of resolver data, and a gateway which can redirect OpenURLs to registered resolvers based on the requester’s IP address.  If both the library and website publisher participate in this project, a user searching for information will find an item, click a link to the gateway and be taken directly to an OpenURL resolver maintained by his or her home library.

Currently, COinS has been adopted by a number of websites and applications, including: Wikipedia; WorldCat; WordPress, a blog publishing system; LibX, a Firefox extension that provides direct access to selected libraries’ resources; and Zotero, a Firefox extension that manages bibliographic information from Web resources.  With COinS, we can achieve a complete and smooth interconnection between library catalogs, Web resources, and Web applications.  For example, a user finds a citation in WorldCat, saves it in Zotero, adds it to Wikipedia; the next person sees the citation in Wikipedia, saves it in Zotero, adds it to his blog, and so on. 

Together with COinS, OpenURL holds the promise of wider, easier access to library resources from various Web resources and applications.  However, as Chudnov reminds us, people “want stuff, not meta-stuff.”  Can people simply re-use library resources within new Web applications?  That is, can users copy items they see online and paste them into desktop applications or other Web applications such as blogs and photo-sharing services?  unAPI provides a method for copying rich digital objects out of any Web application.  It is a tiny HTTP API, application programming interface, for the few basic operations necessary to copy discrete, identified content from any kind of Web application.  A direct benefit of employing unAPI on a website is that it allows other Web users to easily take a piece of its content to create new resources.  In Chudnov’s words, “You see stuff, you get stuff, and you pick the format.”  The unAPI specification is less than two pages and requires very few changes in Web templates.  It can be added to all library resources such as the OPAC, institutional repository, journals, metasearch, and link resolver.  Currently, two major applications using unAPI are Zotero and WordPress.  Chudnov hopes that more website publishers will adopt this new specification.

The next frontier in information services is service links.  Examples are the set of buttons next to an article in a journal or major media websites.  They permit users to email, save, print, and cite in various formats, or send to a bookmarking application such as del.ici.ous.  Libraries can use OpenURL to facilitate this kind of service. OpenURL with COinS can provide user-generated service coordination, and unAPI allows users to choose various formats of the same object.  Chudnov proposed a new specification nicknamed SLAPI, Service Links API, which will fully integrate library resources with free Web resources at the user’s end. 

Finally, Chudnov explored how libraries can work like iTunes, letting users find their friends’ libraries.  One approach is to mesh metasearch and link resolvers, since they work similarly from the user’s perspective.  OpenSearch, a collection of Web technologies that allows publishing of search results in standard format, can further simplify the search process.  When a user’s Web browser knows where a user wants to search and resolve, coupled with SLAPI, a user can access his institutional resources anywhere on the Web, from citation directly into full text.  With Zero Configuration Networking technology, the user will not even need to configure his browser. 

In this ideal environment, everyone visiting your network automatically finds your search/resolver interface, and everyone else you visit finds your institutional resolver. Furthermore, no installation is required on the user’s part.  This full circle, coupled with SLAPI, contends Chudnov, is a new approach to service discovery and resource delivery.