Learning Management Systems: Are We There Yet?

• 06/29/04

The LMS has reached maturity but is racing to keep up with changing requirements on campus and to meet the need for integration with other enterprise systems and a more collaborative working environment. Reminiscent of the kids in the back of the car on your family’s summer vacation, the persistent question about this technology seems to be, “Are we there yet?” Syllabus asked the Andrew W. Mellon Foundation’s vice president for research in information technology about his vision for learning management systems and related technologies.

Syllabus: Where are learning management systems on Mellon’s radar screen, and what kind of priority do they have?

Ira Fuchs Learning management systems have been high on Mellon’s radar screen ever since we made a grant to MIT in 2000 to support the OKI project. The purpose of OKI was to create a new framework to facilitate collaborative development of the components that comprise a modern learning management system.

That eventual goal is still in sharp contrast with where we are today. Now, if an institution acquires a commercial, proprietary LMS, and then finds that the system is deficient in some way, they often must wait until the vendor decides it is financially viable to develop the enhancement—an event that may never occur. Ideally what we’re seeking is a situation in which the schools that want a new capability added to an LMS can, if they wish, develop it themselves, and then make it available to the higher education community so that others may benefit. That’s the point of leveraging collaboration among institutions.

OKI focused on this framework and the delivery of a proof of concept, meaning a system or a pair of systems that could demonstrate this interoperability. And that’s in fact what MIT and Stanford achieved.

S: So OKI focused on the framework… how d'es the Sakai project build on that?

IHF: The Sakai project starts out where OKI left off by taking the architecture and the OSIDs [Open Services Interface Definitions] and fusing them with the best of breed development—learning management system development—from four major institutions: Stanford, MIT, Indiana University, and the University of Michigan. The purpose is to create a world-class production-ready system that will be open, extensible, and scalable. And, further, a very important aspect of Sakai is that the four institutions have agreed, in writing, as a condition of the grant, that they will bring this new system into production on each of their campuses at the same time, approximately a year from now. The goal is really nothing less than delivering an LMS that colleges and universities can use and extend with modules written at other schools, at their own school, or licensed from commercial vendors.

S: Do you think learning management systems will be considered a core technology for colleges and universities going forward? And will open, interoperable systems prevail and be in common use? Are we there yet?

IHF:I think learning management systems are a core technology already, and that fact is, I think, both good and bad. It’s good because learning management systems have helped the faculty and students enormously. They make course information and content available on the Web, and at the same time improve communication among students and faculty. But because the LMS is already so important to the functioning of many schools, it’s going to be hard to move away from the proprietary systems they may be running today and to begin using open, collaboratively developed and maintained systems. I think open systems are going to prevail, but it’s going to take time.

S: So, in a sense, we’re not really there yet…What are some of the steps that could move all of this forward?

IHF: That’s true, we’re not there yet. But Sakai is about to deliver a beta release. The concept is to leverage the work of many, many institutions to ultimately build a system that most, if not all, institutions will want to run. But that’s not the case yet. Today, you have a plethora of choices among learning management systems. There are sites on the Web listing dozens of them. But for institutions seeking to move away from their current LMS, there is a cost to change. The cost comes in many forms, not the least of which is that people grow accustomed to an interface. And often they’ve converted content to be used in that system. So whatever we come up with is going to have to account for and minimize those costs of change.

One way to minimize them is, for example, in the case of the user interface, to have what are commonly known as skins. These are modifiable user interfaces that are selectable by an institution, or sometimes even by the end user, to make the system look the way they want it to look. We’re also going to need to have tools to facilitate the transformation of content from one system to another, to export it and then import it into another system. So we’re going to have to do what we can to minimize the cost of converting from one system to another.

S: Is interoperability among installed systems a key goal for OKI?

IHF: Absolutely, that’s what OKI is all about. The basis for all of this is to have a set of standards, of common interfaces, APIs or OSIDs. I think this is the right time, because people have learned, first of all, that it’s too expensive to try to develop it all on their own. Even the biggest institutions—such as Michigan, the Indiana University, Stanford, and MIT—have decided that building and maintaining these complex systems on their own just d'esn’t make sense any more. At the same time, the notable, visible success of some of the open source projects—the big ones like Linux, Apache, or MySQL—have proven that it’s possible to develop something in the open and get people to commit to maintain and enhance the software.

Perhaps the most important fact to remember is that the industry we represent, higher education, is unique in our willingness to collaborate and to share our labors, such as we have in this IT space. There are a lot of smart people in each of these institutions, and if we can harness them behind the same projects and use a set of standards, starting off with a good base piece of software such as I think Sakai will deliver, then we can do wonders.

S: What about standards for metadata? Is that something to consider along with the interface standards?

IHF: Sure it is, and that is something, of course, that the library community has been working on for a long time. What did someone once say?: “The wonderful thing about standards is there are always so many to choose from…” And we do have many metadata standards. But I think that they will converge, at least in limited domains. When it comes to learning object repositories, it’s going to lead to a set of metadata schema, metadata standards that will not satisfy everyone—that’s probably impossible—but will be good enough. Many of the Mellon-funded projects—OCW, Sakai, LionShare at Penn State, Chandler—are all trying to converge on a common standard for metadata.

S: Will learning management systems change significantly in the next few years? Have they been on the right track, and are they flexible enough to be used universally?

IHF: Learning management systems have come a long way, but there’s still much that can be done to improve usability in particular, especially to make it easier to publish or create new material. It still takes too much expertise to create attractive materials from the notes, images, and programs that faculty use to teach a course. The proliferation of learning management systems suggests that no one system is sufficiently feature-rich, or adequately flexible and extensible enough to meet everyone’s needs or even most institutions’ requirements. But I hope to see that change in the next couple of years with the advent of Sakai.

S: Are new development tools needed?

IHF: Yes, I think we need authoring tools that lower the effort threshold dramatically for faculty to take digitized materials and create something esthetically pleasing as well as effective for their teaching purposes. There are tools, but we have to make sure that they are going to be compatible with all of the other pieces that we’re putting together based on standards. Of course, they’re not yet very compatible, but how could they be? They were built at some point in the past when people weren’t worried about that.

S: What are the pieces needed so that learning management systems can become more easily or better integrated with other parts of the campus information system, either on the academic or on the administrative side?

IHF: We need the middleware layer that translates the standards, such as the OSIDs, for the actual campus infrastructures. For example, OKI defines a set of OSIDs for authentication and authorization, and we want developers to be able to use those OSIDs, so that the systems will be interoperable. However, just about every campus has some authentication system already in place, whether it’s User ID/Password, or Kerberos, or Shibboleth. So there needs to be code which translates the calls that use the OSIDs, to the actual campus mechanisms. This is kind of a chicken-and-egg problem. Why create the middleware unless developers are using the standards? Why should developers use the standards unless the systems they are writing for have implemented the necessary middleware? But I think it’s going to happen.

S: How do portals fit in with all of this?

IHF: There’s another project, which was funded by the Mellon Foundation at almost the same time as OKI that has been very, very successful—that’s uPortal. It’s in use at scores of institutions now. It is the primary enterprise portal at those institutions. So when you ask the question about how to make it easier to integrate the LMS with other parts of the campus information system, I think uPortal is going to play an important role—and Sakai is built on top of uPortal.

S: Will libraries become better integrated with the LMS?

IHF: I think they must become better integrated in-so-far as making it as transparent as possible to the end user—faculty or the student—as to where the information used by the LMS is coming from or how to search for it. And that’s a significant challenge since there are many potential sources for the data used in an LMS. A course can use data from online publishers, from the campus library, from another library, from the campus repository, or even from the faculty member’s local or server-based files. With the emergence of peer-to-peer tools, such as LionShare, the data could even come from the personal machines of individuals throughout the world. Somehow we need to make all of this distributed information available in the learning management system without the user having to learn so many different interfaces.

S: What do you see happening with the development of discipline-specific repositories? Will that be in the purview of the libraries?

IHF: I think so. Repositories that house the scholarly output of an institution are starting to be developed, using software such as DSpace and Fedora, another two projects receiving support from the Mellon Foundation. These are not discipline-specific initiatives, but it’s likely that these repositories will be tied together using the Open Archives Initiative protocol for metadata harvesting, OAI-PMH, to create new sources of published materials. That could in turn have a very important impact on the existing publishing model for scholarly output.

S: What is the status of learning objects? Will they be widely adopted? What kind of timeline would you predict for a widespread adoption?

IHF: Widespread use and adoption of learning objects is one of those things that has been predicted to be, I think the three-letter acronym is RSN—“Real soon now,” for a long time. All I can say is that technology is making it easier to package, search for, and utilize learning objects. But whether we will ever completely overcome the NIH, “Not invented here,” obstacle, only time will tell. The early reviews of the OpenCourseWare initiative at MIT are encouraging and suggest that people do want to be able to borrow, share, and learn from other institutions’ learning materials. For decades, one of the obstacles has been that while faculty may be willing to use another’s textbook or chapter, it’s been very hard to get them to use learning materials, because, after all, teaching is an intensely personal activity. But as we figure out how to package the object, as OCW is doing, I think we’ll see much more sharing than we have in the past.

Ira Fuchs

Ira Fuchs joined the Mellon Foundation in July 2000, in the newly created position of vice president for research in information technology. He is responsible for directing the foundation’s expanding investigations of digital technologies that can be applied to teaching and research. Fuchs is also senior technology advisor to the president of Princeton University. From 1985 to 2000, he was vice president for computing and information technology at Princeton. In 1981 he founded the BITNET Network, the first and world’s largest academic telecommunications network, later serving as president of its successor, the Corporation for Research and Educational Networking (CREN). He continues to serve as a trustee of JSTOR, a rapidly growing digital archive of more than 400 core scholarly journals.

S: Will learning objects have an impact on commercial publishers, especially textbook publishers, and how will materials like electronic course packs be used with the learning management systems?

IHF: I envision textbooks that utilize technology in ways that go far beyond merely including a CD with a searchable text. I can imagine textbooks that are distributed digitally, read on a handheld reader with screens that can rival paper’s readability. Sony recently announced a handheld device with a 200-pixel resolution instead of the 72-pixel resolution found on a standard monitor. That really starts to look like paper. And as storage capacities increase, you’ll get more than the text of the textbook. You’ll have videos of lectures. You’ll have supporting materials. The LMS of the future will be a distribution point for the digital textbook. Electronic course packs, I think, will also be distributed via the LMS.

S: Are learning technologies and new media in synch? Can learning management systems incorporate new media such as streaming video effectively?

IHF: I think future learning management systems will make use of both one-way and two-way video. In terms of one-way video we are starting to see streamed video, for example, lectures in OCW for a limited number of courses, and they’re very popular. At the Harvard Medical School, they routinely video all lectures in the first two years of study and make them available for viewing immediately after they are recorded. I hope that many of these recorded lectures are going to be made available freely on the Web in OCW-like repositories—a goal of OCW is to facilitate and foster at other campuses what they’ve done at MIT. I also think two-way video in the form of desktop video conferencing is going to become much more prevalent as the technology improves. However, one prerequisite for the effective use of two-way video is better end-to-end bandwidth and guaranteed quality of service (QoS)—both unfortunately are not common today on the Internet.

S: In terms of the commercial course management systems, how will they have to change in order to continue in that market? And is there room in the market for more players?

IHF: This a question that gets asked all the time, in relation to Sakai. With OKI the answer to questions about competition with commerical offerings was that we weren’t really building a production system, only a proof of concept and a framework. But with Sakai it’s no longer ambiguous. It’s clear that the intent is to build a production-quality, open source, open standard, extensible system. That’s what we want to do. And, so, I’ll say what I’ve been saying for a long time. It’s my hope that proprietary course management systems will, over time, be supplanted by open systems that facilitate collaboration and sharing of both course materials and content, with software designed to extend the CMS itself.

I think there’s still going to be plenty of room for commercial firms to develop and sell modules that fit into the open framework. I also think there will be a significant demand for services related to course management systems, including maintenance and custom enhancements. However, a new business model will be required. The model we have today is that you buy a software black box and you get tied to it. If you want extensions, you hope that they’ll be in a future release. Further, there’s no ability to leverage the brilliant minds on all our campuses to make these systems better. We want a different model, one in which we can collaborate among our campuses on improvements. I think there will continue to be opportunities for others who have a for-profit model to build a better mousetrap—provided it fits into the framework that we’ve developed so that we don’t have to buy the whole package to get one piece that we like.

S: You’ve talked about a lot of open source technology that is related to learning management systems. How will all of this—Sakai, LionShare, Chandler —be supported in the future?

IHF: Lack of reliable support is the issue that is mentioned most often as an obstacle to the widespread adoption of open source software in higher education. Let’s look at some of the major open source, higher education-related projects underway today and see where those projects may be leading us.

Consider the Chandler project, an initiative to create the next generation of personal information management tools, including enhanced e-mail, calendaring, and flexible links to all of the information we use on a daily basis. Chandler ’s goals are ambitious but what really makes this project special is that it is being developed by the Open Source Applications Foundation (OSAF) led by Mitch Kapor, a software entrepreneur with an impressive track record of delivering quality software, and it is being supported by 25 major research universities—the members of the Common Solutions Group, each putting up $50,000. Their contributions are further leveraged by matching funds from the Mellon Foundation and the OSAF itself for a total of over $6 million in funding. In other words, these schools are outsourcing the development of open source software.

Then consider the Sakai project. Beyond the $2.4 million grant from Mellon, the four major institutions are collectively making an in-kind contribution of over $4 million in the form of staff that they are dedicating to the project and placing under the management of a joint board. In addition, they have created the Sakai Educational Partners Program (SEPP) and they have invited schools interested in the future of learning management systems and in particular the development of Sakai to participate in the project. Partner institutions contribute $10,000 per year over three years and in return receive early releases and significant input into the development of the system. Approximately fifty schools have signed up so far and we’re hoping that as many as two hundred will join.

So, where is this leading? Assuming the success of these projects, I think it’s all leading to the creation of an organization that will coordinate and provide funding for open source software projects to benefit higher education. I’ve given this organization the working name of Educore, similar in some ways to Bellcore, an organization that was created to serve the Baby Bells. I think people will see it as a natural next step. Looking at the success of Sakai, they may say, if you could do that, then why aren’t you doing more? In a sense, we’ve got a set of dots, representing OKI, Sakai, LionShare, Chandler, and some of the other projects. Draw a line through the dots, and where d'es it lead? I think it leads to doing more in the same open, collaborative way, with a set of standards that ties all the pieces together. I think that Educore would be the natural progression, moving up one order of magnitude, from the series of initiatives that are underway today. If one to two thousand schools were each to contribute an average of $10,000—which is, by the way, less than they probably spend on one proprietary software package—you could then create an organization that could support 100-200 FTEs. With that level of funding, I believe it would be possible to attract the talent that would be necessary to both develop and maintain the key IT software needed by higher education including learning management systems, authorization/ authentication middleware, personal information management tools, and more.

S: What would the governance be? How will this be organized?

IHF: The question of governance is very important and would certainly have to be addressed, along with many other issues such as how to set Educore’s priorities. There are also many models for open source software development in use today and Educore could coordinate projects that range from supporting lone developers at a university, to outsourcing to large centralized groups such as the one at OSAF or OSDL, to doing the work in-house. Educore would undoubtedly present many challenges but I think it is doable and of potentially enormous value to higher education.