Hello harry, sorry for the delay. As explained in the post, we can use the DBSCAN algorithm to create clusters of users based on their similarity and then recommend products between them. This is based on an approach called collaborative filtering which is a well-known approach in recommender systems.