What are the latest visions, problems and topics of your Institute's new developments?

Yael Ravin: It is well known that 70-80% of learning happens "on the job", through collaboration with colleagues or by solving problems while working, but 80% of the investment in learning is still targeted towards formal learning -classroom training, LMS, or virtual classrooms. It is important to continue to develop formal learning, but there is a real need for tools and processes that can help this informal learning. An important goal at the IAL is to address this need and develop tools people can use on the job or while working.

The first thing we did was to start with a definition of "informal learning" - what is it? We developed two models of informal learning. I'll talk about these models in the context of corporate work, but they are also applicable to academic learning situations.

The first model is Enabled Learning. In this model a person has a specific, immediate, learning need. For example they need to brush up on a new product or learn more about an industry trend in preparation for a customer visit the next day. There is no time to take a course. In this case, Enabled Learning can provide a short guiding experience, tailored to the specific need, available immediately, within the context of work. This can be a "mini-course- assembled dynamically in response to the person's query or a short collaborative session enabled by locating a relevant human expert.

The second model is Embedded Learning. In this model learning is intimately embedded in the work process itself. For example a person is about to execute the next step in a workflow application. If they are not sure how to proceed, modular learning about this specific step is invoked to bring the person up to speed quickly. Or a chat with a human expert who can guide the user can also be initiated. In this case Embedded Learning allows the user to accomplish their work by providing learning at the moment of task execution.

Our challenge is to develop tools, technologies, and methodologies to facilitate both enabled and embedded learning.

What was its biggest success during the last two years?


Yael Ravin: We can point to two recent successes at the IAL - two systems we developed: Dynamic Learning Experience (DLE) for the corporate learning environment and Teacher Workplace for compulsory education teachers. I'll describe our work on DLE, which addressed the Enabled Learning need we talked about. Users input a query (e.g., "Web Services") and their preferences (how much time they have, whether they'd like an overview or in-depth study, the level of difficulty, etc.). DLE dynamically assembles courses out of a repository of learning objects, respecting the user preferences.

The system arranges the objects in a logical order in accordance with basic rhetorical and instructional principles and delivers the custom course to the user. We conducted several pilots of DLE inside IBM with our technical population and have now deployed the system internally to hundreds of technical consultants. More recently we started a similar pilot with the Advanced Distributed Learning Initiative (the ADL) of the US Department of Defense.

How much time is needed to come from vision to applicable technology in learning questions?

Yael Ravin: The answer varies. For DLE it took about three years from the very early idea in the mind of the researcher to the current implementation with a client. There was, however, continuous progress. Early and continued collaboration during the first pilots between the IAL and the Learning organization of IBM Global Services was essential to our success.

We had many opportunities to get formal and informal feedback from a wide range of users, and we incorporated many changes to the design as a result. We also conducted a formal experiment comparing users' performance on a task using DLE with a control group that used a regular search engine. The results were very positive.

At which point does the IBM Institute for Advanced Learning step out of the process?

Yael Ravin: The IAL stays in the picture until the asset is successfully transferred to an "owner" in IBM. With Teacher Workplace (TWP), for example, we are at the end of this process. We started three years ago with brainstorming sessions with teachers to determine what tools would help them in carrying out their daily tasks and improve student performance. We then developed the system iteratively through pilots with several school districts and Education colleges. When the first prototype version of TWP was stable, we transferred the technical development to an IBM group outside of the IAL. They are now working on the second release of TWP.

They are also providing support to the users at the pilot sites. We even engaged the education consultants at IBM to help them develop a business plan for marketing and selling the asset and consulting services to help school districts improve student performance. We continue to provide occasional support, either technical or by discussing the technology with clients, but we see our role diminishing and are now looking to start our next education project.

The IBM researchers are working in New York, San Jose, and Haifa, Israel. What kind of virtual project management, new ways of communication, and technology-based learning forms have they needed and used to handle their tasks?

Yael Ravin: There are groups at IBM who work very closely around the globe. A lot more can be done to help them. They still rely very heavily on the phone. We all do. There are days when I don't have time to go to the office because I am on conference calls from seven in the morning till six in the evening. IBM Research is developing a variety of new communication tools, including one to make conference calls easier and another to manage and facilitate collaborative code development.

The IAL work, however, is not so tightly integrated. So the group in Israel, for example, has world-leading expertise in search technology. We leveraged this in New York and used the search engine from Israel as a component of DLE. For this reuse, a couple of trips and e-mail were sufficient. But this is not always the case.