It is used to identify patterns, relationships, or trends that one may have been previously unaware of in a dataset.