Dictionary.net

An algorithm for encoding a word so thatsimilar sounding words encode the same. The first letter iscopied unchanged then subsequent letters are encoded asfollows:bfpv -> "1"cgjkqsxz -> "2"dt -> "3"l -> "4"mn -> "5"r -> "6"Other characters are ignored and repeated characters areencoded as though they were a single character. Encodingstops when the resulting string is four characters long,adding trailing "0"s if it is shorter. For example, "SMITH"or "SMYTHE" would both be encoded as "S530".