Regresar

Genetic algorithms for courseware engineering

Abstract:

The process for creating learning contents using reusable learning objects (LOs) can be broken down into two sub-processes: LO finding and LO sequencing. Finding can be automated by the use of federated search engines along with gap analysis techniques; however, sequencing is usually performed by instructors, who create courses targeting generic profiles rather than personalised materials. This paper proposes a novel technique for courseware engineering that aims to solve two recurrent problems in this area: automation of the instructor's role and personalised course building. Simultaneously, e-learning standards are promoted in order to ensure interoperability. Competencies and metadata are used to define relations between LOs so that the sequencing problem turns into a constraint satisfaction problem and a genetic algorithm is designed and implemented to solve the sequencing problem. The proposed agent is tested in real and simulated scenarios. Results show that it succeeds in all test cases and that it handles reasonably the computational complexity inherent in this kind of problem. © 2011.