In the domain repository, similarity of domains could be determined in order to obtain better and more precise decisions.
Furthermore, this information could be updated after each decision making process with respect to new results in various domains.