Phonetic Matching is the process of matching data using Algorithms / functions that have been created focusing on how a word is pronounced rather than how it is spelt. Most of the popular phonetic matching algorithms are designed to work with the English language and do not perform as well with foreign languages.
The other common approach is Edit Distance, which looks at the differences in strings rather than pronunciation, so this picks up typos rather than misspellings. Phonetic Algorithms are commonly used in spell checker applications.
Popular Phonetic Functions are:
Phonetic matching is particularly suited to matching names where the different spellings are used but the pronunciation is the same for example, John, Jon or Clare, Claire.
Phonetic Matching is not helpful in dealing with typographical errors where the keys have been pressed in the wrong order for example ‘Andrew’ instead of ‘Anrdew’.
Its also far to say that whilst Phonetic Matching is very helpful in some instances in is less so in others for example in Matching Company Names where the company name is made up of multiple words rather than a single word which is typical of the majority of English first names.