Although traditional intelligent tutoring systems have successfully fostered learning, few systems have been built due to high content creation costs. Previous research on the ASSISTment system reported a reduction in these costs through the use of pseudo-tutors. However, building isomorphic content - problems that are identical except for surface features - was expensive because each problem had to be built independently. In this work we present novel features of the ASSISTment system that allow rapid creation of isomorphic content via templates. Through a range of experiments and case studies, we report evidence that templates reduce time required to build isomorphic content, do not decrease ease-of-use of the system, and can be used to develop personalized and culturally-aware tutors.
Worcester Polytechnic Institute
Interactive Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.