Genetic algorithm for automatic group formation considering student's learning styles

Group formation is an important topic in Computer Supported Collaborative Learning (CSCL) because that has implications in the group performance. In this paper, we propose a genetic algorithm for automatic generation of groups considering learning styles of your members. The group formation with genetic algorithm is a permutative problem, for this reason, genetic operators were designed. We use historical data about performance of groups and we create association rules which are used in the fitness function. The algorithm proposed was analyzed with different size of groups given for the teacher. Through the experimentation we can see what kind of configuration tends to be more appropriate.