Assisting analysts in the construction of model based tracking recommender systems

Model based tracking is an approach that allows the construction of recommender systems based on the comparison of the user´s navigation with predefined prototypic models of navigation. This approach is well adapted for complex Web sites and allows overcoming a certain number of drawbacks from classical approaches. However, it requires an additional task: analysts must build the reference models and their associated tips. In this paper, we present an approach and tools to support the analysts in this task.