Intelligent Algorithm Development Tutor
Middle Tennessee State University, Murfreesboro TN
Investigators
Abstract
Computer Science (31) This project is developing a pedagogical tool that trains introductory computer-programming students in using top-down design with abstraction refinement. The tool also allows students to use the bottom-up approach to utilize previously solved modules. This system helps students to develop sound problem solving skills at the earliest stages of training. Specifically, this online tutoring system provides a personalized assistant for understanding the algorithm refinement process. Students can reinforce these problem solving skills by using the tutor to solve a number of specific problems. The proposed project extends a previously developed online tutoring system for laboratories for introductory computer science courses to include tutoring for top-down design concepts. Solutions to selected problems are represented at multiple levels of abstraction. Alternative solutions are represented, again at varying levels of abstraction, as are "dead-ends" and other erroneous solutions. Solutions to related problems are also encoded. Collectively, this precompiled knowledge forms an abstraction network. The abstraction network allows students the flexibility to design their own solutions to problems. In addition, it permits students to follow faulty paths so that they can discover the implications of earlier design choices and learn from their mistakes.
View original record on NSF Award Search →