This is a broader question - but do you have any insight on how to convert e.g. Levenshtein distance to probability? I want to combine the edit distance with prior information on the background population - and it's not clear to me how to combine the two metrics. Thanks!