Can this algorithm combine with other, such as Levenshtein Distance, to get get distance of Chinese phrases of the different length?