Learning Systems & Us

Not so long ago, when we talked about e-learning, people thought of it as technology to support distance learning. Now that technology goes beyond distance learning, and many view it as a standard resource for "bricks and mortar" classes.

Software from companies like Blackboard and WebCT now support common class needs such as syllabi, quizzes, discussions, and assignments. In the most recent Educause Core Data Service (from 2002), nearly 92% of the responding institutions had a course management system, also called virtual learning environment or learning management system. The same survey revealed that nearly 89% of the institutions had a library management system. Clearly, the popularity of learning management systems is no longer driven solely by distance education needs.

As recently as two years ago, adding library content to a course in a learning management system wasn't on the radar of many, including faculty. Several learning management system vendors made it possible to license content in conjunction with their software—frequently content the library already licensed. Fortunately, there is now greater recognition that the library has the richest content repository on campus and needs to be integrated into the learning management system.

Recent developments in library technology dovetail well with this need to integrate library content into learning management systems. Both OpenURL resolvers and federated search products are important tools to help accomplish this goal. The OpenURL standard represents a method to incorporate a locator, or a persistent way to find the document, such as class readings. The federated search—a single interface that allows a user to send the same search to many resources, then view all the results—can help the learner dynamically discover new information.

The OpenURL resolver plays a dual role. Primarily it should create a durable link to the resource in an external system, such as from a citation in an abstracting and indexing database to the full text of the article. Today, frequently, the reading list in the learning management system does not link to the actual article, forcing the student to take extra steps (opening another browser, going to the library site, and searching for an e-version). In some cases, faculty may have found the article on a vendor's site and cut-and-pasted the URL to their reading list. Often, later, the link may not work because of session-specific information in the URL.

Associating OpenURLs with reading lists will help solve this problem. By creating an OpenURL for the citation, the user is taken to the resolver for a link to the full text of an article, or if supported by the resolver, the OpenURL is automatically resolved and the user bypasses the resolver menu. For those citations available electronically through a local e-reserves system, the OpenURL could use standard metadata to access the course reserve system or OPAC.

The other role for the OpenURL resolver is creating or pushing information into the learning management system for resource lists like required course readings. Typically, the course creator finds citations in a variety of citation databases or from a publisher's e-journal site. Generally, there has been no easy way to pass an OpenURL from a resolver to a particular course for inclusion in a reading list. But librarians have developed creative solutions, such as providing the OpenURL for copying into a course. Without programming effort, that is the limit to what librarians can do.

Help may be on the way. Most of the learning management systems provide an open set of Application Programming Interfaces (APIs) to enable third-party application integration. APIs are a documented way for applications to exchange data directly. Unfortunately, there are no standards in this area, and the APIs are all proprietary. Library vendors must develop integration with each learning management vendor separately. Typically, vendors try to make their APIs backward compatible, but changes sometimes mean that a third-party application can no longer interoperate—especially for major releases. A developing standard promises a single approach to integrate citations.

As the role of learning management systems expanded, the e-learning community understood the need to interact with a variety of systems, such as student information systems and authentication services. The IMS Global Learning Consortium was formed to develop and promote standards to enhance interoperability among various systems.

An IMS working group is developing a standard to move lists of bibliographic descriptions of learning objects among systems. A first draft of the standard has been released (www.imsglobal.org/rli).

Such interoperability offers interesting possibilities for integration between library systems and learning systems. In the case of the OpenURL resolvers, it may be a solution for more seamlessly passing back a citation and its OpenURL to the learning management system. In this scenario the faculty member would find an article in a citation database, click on the OpenURL associated with it, and display the resolver menu. The menu would include an option to transfer the citation to an existing course in a learning management system. By selecting that option the metadata associated with the OpenURL and even the OpenURL itself are passed to an LMS using the protocol defined by the resource list interoperability proposal.

But the opportunity for interaction with library software goes beyond OpenURL resolvers. It can be used to pass a resource list from the learning system directly to an integrated library system for management within the reserves module or to a standalone reserves system. This would be a vast improvement over current ad hoc solutions. By using a standard metadata format some automatic processing will be possible, e.g., using citation data to search the local catalog, match citations, and generate pick lists for print items.

While managing reading lists is important, students are frequently given assignments or group projects that require research. Increasingly, librarians are concerned that their students are turning to Internet search engines, such as Google, first. An easy-to-use search mechanism must be available within the learning management system, where, after all, students spend so much of their time.

Federated search products are ideally suited to help students, and the single search interface has proven popular with undergraduates. Most promising: the course creator can harness the benefits of the federated search engine and tailor searching for a particular course. There are three primary scenarios for integrating a federated search product into a learning management system: the predefined search, an embedded box search, and the embedded search interface.

With the predefined search, the instructor creates a search in the federated search product using the most appropriate resources for that topic. The student simply sees a link in the course, clicks on it, and the predefined search automatically executes.

When the search box is embedded, the course content creator places a search box in the course and associates the most appropriate resources to search. The student sees a search box within the course itself and simply enters a search string and clicks the search button to link to the resources. When the complete user interface for the federated search product is displayed within the context of the learning management system, all the standard features of both products are available.

In each scenario, the search results are either spawned in a separate browser window or viewed framed in the interface of the learning management system. In all cases, the student still has access to the familiar course interface.

A critical feature to help the course instructor provide this level of integration is being able to add library content as easily as other standard course components such a quizzes, course assignments, etc. To provide such integration the federated search vendor must rely on the proprietary APIs provided by the learning management system vendor. Given the critical importance of adding the content and difficulty of ad hoc solutions, the vendors are willing to take on this work despite the issues mentioned above.

At this point, the learning management systems are willing to have the searching done by third-party applications. The barriers to providing federated searching are high since the learning management system would need to translate a search into various protocols and understand and display wildly divergent metadata formats such as MARC, HTML, and XML-based records.

One of the models being considered by the Metasearch Initiative of the National Information Standards Organization may bring relief. Under discussion is an application, such as a learning management system, that uses a standard protocol to send a search to a federated search engine that is, essentially, a broker. The federated search engine then sends the search to requested resources in the correct format. This would make the job of the federated search engine easier, since it would no longer be necessary to write to various proprietary APIs.

It is critical that the library partner with instructional designers as well as the key faculty who have embraced the use of learning management systems. The campus units that support the learning management system are frequently unfamiliar with library content.

Another critical activity is to let library software vendors know the value of integration with the learning management system. Obviously, as librarians look at OpenURL resolvers and federated search engines, they will want to make sure that e-learning integration is on the vendor's product roadmap.

Finally, it is important to educate the library staff about the learning management system. Find out how faculty are using it. Is it primarily to post the syllabus and reading lists or is there more substantial activity taking place? One of the best ways to understand the product is to deliver bibliographic instruction through a learning management system.

The last 12 months have shown significant progress in integrating library content into learning management systems. There are still significant issues to address, but it is encouraging to see the communities talking about interoperability. Perhaps more important is that the educators are now asking for our content. Together we can drive additional development.