The recommender systems are our interfaces with the vastness of the Internet, and their role is crucial in opening new doors and proposing new content to us and they have been optimised to stay as close as possible to users’ preferences.

Starting from this point, our research work aims at building a new generation of algorithms capable of showing us new paths, without pushing us into unpleasant or incomprehensible experiences. On the other hand, we are developing a testing framework and metrics to study and forecast the effect of algorithmic guided exploration on the opinion dynamics.