Re: Search for similar strings - to standardise

[Rival]

in reply to Search for similar strings - to standardise

For the string comparison, using the Levenshtein distance is probably your best bet. This is often used in spell-checking to suggest corrections.

For implementations, there is a Text::Levenshtein module on CPAN. But also, take a look at this node and this node.

[educated_foo]

Edit distance would be useful for comparison, but not so much for clustering strings with their common misspellings. For that, you might try n-grams (mentioned above) or locality-sensitive hashing (basically the same thing, but with gaps).