I use the levenshtein distance for NLP data cleaning/spell correction. I also use a bloom filter that can validate if the word does not exists in my corpus, so

I use the levenshtein distance for NLP data cleaning/spell correction. I also use a bloom filter that can validate if the word does not exists in my corpus, so I don't have to run the algorithm for every word.