In this article, Luis Rivas describes another way to leverage this power, and that's the discovery of custom category detection including variants (misspellings or non-standard category names) and automated data standardization or grouping of these detected categories.