This involves identifying patterns in a large dataset by grouping different variables that are related to each other.