Common Characters: Find and group values that have letters or numbers in common. This option uses the ngram fingerprint algorithm that indexes words by their unique characters after removing punctuation, duplicates, and whitespace. This algorithm works for any supported language. This option isn't available for data roles. For example, this algorithm would match names that are represented as "John Smith" and "Smith, John" because they both generate the key "hijmnost". Since this algorithm doesn't consider pronunciation, the value "Tom Jhinois" would have the same key "hijmnost" and would also be included in the group.