Authoring and Sharing of Programming Exercises

In recent years, a number of exercises have been developed and published for educating students in the field of Computer Science. But these exercises exist in their own silos. There is no apparent mechanism to share these exercises among researchers and instructors in an effective and efficient manner. Moreover, the developers of these programming exercises are generally using a proprietary system for automatic submission and grading of these exercises. Each of these systems dictates the persistent format of an exercise that may not be inter-operable with other automatic submission and grading systems. This project provides a solution to this problem by modeling a programming exercise into a Learning Object metadata definition. This metadata definition describes the learning resource in terms of its contents, classifications, lifecycle and several other relevant properties. A learning Object (LO) is persisted in a repository along with its metadata. This repository supports simple and advanced queries to retrieve LO s and export them to various commercially available or home-grown e-learning systems. In a simple query, keywords given by the user are matched against a number of metadata elements whereas an advanced query allows a user to specify values for specific metadata elements