On the design of tele-learning systems using genetic algorithms

Abstract

A tele-learning system may be modeled using an object-oriented approach which lends to eventual simulation using object-oriented languages. In this study, the human users of the system are included as system components with their own respective characteristics, as described by such descriptive social-economic parameters as user satisfaction, user willingness to pay, and others. The current focus of the study is on the genesis of tele-learning systems, that is, a specification of fundamental descriptive parameters, somewhat akin to genes in genetics, that may be encoded as binary strings, and of a fitness function of these parameters which is to be optimized using genetic algorithms. The fitness function measures some critical aspects of the system's behaviour of interest to the system designer. The strings corresponding to the optimal value of this function then will specify an optimal system design.