I have a set of data with first/last name sets from two different sources. In three consecutive records, John Smith might be matched with Mary Smith, John Smith and Jan Smith. It seems like there should be a way to use some kind of fuzzy logic to score the degree of similarity between the sets of name fields, then use that to select the most likely match.
I've gotten as far as using the soundex() function on each name, and can look at the data and see which records are the most likely matches, but I can't find a function that will give me any sort of indicator of similarity between two fields within the same record. Really, I just need a way to recognize that "N265" resembles "N260" more than it does "K357". I could expand the records vertically and use a standard Fuzzy Match, but I'm looking at doing this on a large dataset, and that might be problematic.
