What did the Levenshtein distance look like between the data: how did the algorithm determine a potential match with a record in the sanctions list?