Use Jakarta Commons Codec’s
Soundex. Supply a surname or a word, and
Soundex will produce a phonetic encoding:
// Required import declaration import org.apache.commons.codec.language.Soundex; // Code body Soundex soundex = new Soundex( ); String obrienSoundex = soundex.soundex( "O'Brien" ); String obrianSoundex = soundex.soundex( "O'Brian" ); String obryanSoundex = soundex.soundex( "O'Bryan" ); System.out.println( "O'Brien soundex: " + obrienSoundex ); System.out.println( "O'Brian soundex: " + obrianSoundex ); System.out.println( "O'Bryan soundex: " + obryanSoundex );
This will produce the following output for three similar surnames:
O'Brien soundex: O165 O'Brian soundex: O165 O'Bryan soundex: O165
Soundex.soundex( ) takes a string, preserves the
first letter as a letter code, and proceeds to calculate a code based
on consonants contained in a string. So, names such as
“O’Bryan,”
“O’Brien,” and
“O’Brian,” all
being common variants of the Irish surname, are given the same
encoding: “O165.” The 1 corresponds
to the B, the 6 corresponds to the R, and the 5 corresponds to the N;
vowels are discarded from a string before the
Soundex code is generated.
The Soundex algorithm can be used in a number of
situations, but Soundex is usually associated with
surnames, as the United States historical census records are indexed
using Soundex. In addition to the role
Soundex plays in the census,
Soundex is also used in the health care industry
to index medical records and report statistics to the government. A
system to access individual records should allow a user to search for
a person by the Soundex code of a surname. If a
user types in the name “Boswell” to
search for a patient in a hospital, the search result should include
patients named “Buswell” and
“Baswol;” you can use
Soundex to provide this capability if an
application needs to locate individuals by the sound of a surname.
The Soundex of a word or name can also be used as
a primitive method to find out if two small words rhyme. Commons
Codec contains other phonetic encodings, such as
RefinedSoundex, Metaphone, and
DoubleMetaphone. All of these alternatives solve
similar problems—capturing the phonemes
or sounds contained in a word.
For more information on the Soundex encoding, take
a look at the Dictionary of Algorithms and Data Structures at the
National Institute of Standards and Technology (NIST), http://www.nist.gov/dads/HTML/soundex.html.
There you will find links to a C implementation of the
Soundex algorithm.
For more information about alternatives to Soundex
encoding, read “The Double Metaphone Search
Algorithm” by Lawrence Philips (http://www.cuj.com/documents/s=8038/cuj0006philips/).
Or take a look at one of Lawrence Philips’s original
Metaphone algorithm implementations at http://aspell.sourceforge.net/metaphone/.
Both the Metaphone and Double Metaphone algorithms capture the sound
of an English word; implementations of these algorithms are available
in Commons Codec as Metaphone and
DoubleMetaphone.