|
|
|
|
|
|
| |
Interview with: Cecilia d'Oliveira '77, SM '79Cecilia (Cec) d'Oliveira, technology director for MIT's OpenCourseWare program, earned a master’s degree from the Sloan School of Management in 1979 and an SB in electrical engineering and computer science in 1977. She worked in the MIT Information Systems department from 1983 to 1996, left in 1996 to found SupplyWorks, then returned in September 2002. What were the crucial decisions involved in moving OCW from brute-force HTML coding for the first 32 subjects to a content management system? MIT OCW’s initial Web site opened to the public in September 2002 with content from 32 MIT subjects, with an additional 18 subjects added to the site in December 2002. These 50 subjects, from 23 of MIT’s academic disciplines, constituted the MIT OCW Phase I pilot. The work to develop the pilot sites began early in 2002 before MIT OCW existed as a formal MIT department. Work was done by a couple of staff members working with outside contractors and faculty, and management of the work relied on CVS, a source code control system, and lots of colored Post-it notes. The primary technical tools were Dreamweaver and Photoshop. This is an environment that I refer to as “brute-force HTML” coding. We did what we had to do during the pilot to get the job done and the initial course materials published, but it was clear as we looked forward that the processes and the tools we had used during the pilot phase would not scale. MIT OCW’s goal for September 2003 is to have the course materials from 500 MIT subjects online, a significant increase over the pilot site. In fall 2002, we identified several key requirements that led to our decision to implement a content management system (CMS):
The next crucial decision involved the selection of a CMS software platform. Given our timeline, we had approximately six weeks to decide on a software platform, so we worked from a short list of vendors. We pulled together our high-level requirements and then zeroed in on a small number of vendors, incorporating the input of the consulting firm Gartner and Forrester. Sapient Corporation, our implementation partner, also provided a lot of guidance, given their experience with CMS implementations. We did not include an open-source platform in our short list given our limited technical staffing within MIT OCW and our belief that the existing open-source platforms did not meet our workflow requirements. In the end, we selected the Microsoft CMS 2002 as our target platform. The primary factors in this decision were total project cost, the pre-existing MIT-Microsoft relationship, and ease of use. As the CMS implementation proceeded through the winter, a few of the crucial decisions we faced included the site’s overall information architecture (department, subject, section, resource), the detailed workflow process, the metadata strategy, how to get course materials from faculty into the CMS environment, where to host the CMS, and how to move CMS content to our Unix web servers and our worldwide content distribution environment provided by Akamai Technologies. We completed our initial CMS implementation and began actual production use in April 2003, and we began publishing the MIT OCW web site from the CMS in June. As of July 30, we have 120 courses published from CMS, with an additional 280 in various stages of our quality assurance (QA) process. Between now and the end of September, our plan is to publish an average of 50 courses a week from our CMS -- and by September 30 the Web site will offer the course materials from 500 MIT subjects! What are the benefits of OCW establishing an efficient, standards-based model for other universities to use? MIT OCW has two primary goals as a project. First, we will provide free, searchable, coherent access to all MIT course materials. Second, we will create an efficient standards-based publishing model for delivering course materials. This second goal has both an internal (within MIT) and external (beyond MIT) dimension in terms of benefits. Internally, it is important that we create an efficient publishing model that adds value for MIT students and faculty, without putting undue burden on the faculty. If we don’t do this, faculty will choose not to participate in what is a voluntary initiative, and MIT OCW won’t be sustainable as an MIT activity in the long term. Externally, it is important that we create a publishing model that other universities can emulate to publish their own courses. MIT OCW isn’t just about MIT providing its own course materials to the world. MIT and its sponsors also hope to stimulate other schools and their faculty to make their course materials available, and in the long term, to create a global movement toward free and open distribution of educational materials. As MIT President Charles M. Vest has said, “We hope the idea of openly sharing course materials will propagate throughout many institutions and create a global web of knowledge that will enhance the quality of learning and, therefore, the quality of life worldwide.” We are breaking a lot of new ground and we are learning a lot about the issues in providing access to course materials in a grand scale. As we do this, we will share our best practices with other universities. What technological challenges continue to face MIT OpenCourseWare? First, we need to make sure our content is as accessible as possible to all users -- using Microsoft or Unix, on a cable modem connection or a dialup connection, sighted or visually impaired, etc. Second, we need to put in place the infrastructure to measure the use, access, and impact of our materials so that we can meet our project evaluation goals. Third, we need to evolve our metadata and search capabilities to insure that MIT OCW course materials are as searchable as possible. Finally, we need to develop processes and mechanisms to archive the content of MIT OCW, as that becomes necessary. We will be working closely with the MIT Libraries' DSpace initiative on this.
|
|
|
Copyright ©2003 Massachusetts
Institute of Technology