Ant colony algorithm and new pheromone to adapt units sequence to learners´ profiles

The use of new information and communication technology is increasingly common nowadays. Content adaptation of learner´s profile is an issue that concerns many researchers in education field. Several studies have been conducted to achieve high quality learning and adapt the content to learners´ profiles. Some researchers have properly applied the ant colony algorithm to the field of e-Iearning. In this work we are interested in the improvement of ant colony algorithm for scheduling units of courses (e.g., a Java course). We follow a pedagogical way to establish units. We define five concepts to maintain learners´ motivation and adapt the algorithm behavior to our context. So our contribution is a new pheromone that influences the algorithm to choose the right unit in a pedagogical sequence. Many changes are taken into consideration to implement the new version of ant colony algorithm. The trainers apply weights to each arc that are linking two units of the course. The profile definition is a part that was preliminary defined in previous work using fuzzy logic method [2]. Method Roulette Wee I is applied for the selection part. This method is interested in finding the final state. It is used in addition to the ant colony algorithm for the path exploration and optimal learning path.