Encode words with classic phonetic algorithms — Soundex, Metaphone, NYSIIS, and Caverphone — to find names that sound alike despite spelling differences. Useful for fuzzy name matching, data cleanup, and deduping contact lists.
Soundex (1918, US Census) produces a 4-character code: the first letter plus three digits encoding consonant groups. Vowels and H/W are usually dropped.
Metaphone (1990, Lawrence Philips) produces variable-length codes that better capture English pronunciation, handling digraphs like PH, GH, SCH.
NYSIIS (New York State Identification and Intelligence System) is a fixed-length variant tuned for American surname matching.
Caverphone (2002, New Zealand) was designed to match names regardless of spelling in electoral rolls.