Jaro-Winkler distance

The Jaro-Winkler distance (Winkler, 1999) is a measure of similarity between two strings. It is a variant of the Jaro distance metric (Jaro, 1989, 1995) and mainly used in the area of record linkage (duplicate detection). The higher the Jaro-Winkler distance for two strings is, the more similar the strings are. The Jaro-Winkler distance metric is designed and best suited for short strings such as person names. The score is normalized such that 0 equates to no similarity and 1 is an exact match.

The Jaro distance metric states that given two strings $s\_1$ and $s\_2$, their distance $d\_j$ is:

:$d\_j\; =\; frac\{1\}\{3\}left(frac\{m\}$

* $m\; =\; 4$ Note that the two "X"s are not considered matches because they are outside the match window of 3.
* $|s\_1|\; =\; 5$
* $|s\_2|\; =\; 8$
* $t\; =\; 0$

We find a Jaro score of:

:$d\_j\; =\; frac\{1\}\{3\}left(frac\{4\}\{5\}\; +\; frac\{4\}\{8\}\; +\; frac\{4-0\}\{4\}\; ight)\; =\; 0.767$

To find the Jaro-Winkler score using the standard weight $d\; =\; 0.1$, we continue to find:

* $ell\; =\; 2$

Thus:

:$d\_w\; =\; 0.767\; +\; (2\; *\; 0.1\; (1\; -\; 0.767))\; =\; 0.813$

References

*
*
*
*

ee also

record linkage, census

External links

* [http://www.dcs.shef.ac.uk/~sam/stringmetrics.html#jaro Open Source implementation in Java and .NET]
* [http://www.census.gov/geo/msb/stand/strcmp.c Original C Implementation by the author of the algorithm]
* [http://diotalevi.isa-geek.net/~josh/Jaro-Winkler/winkler's.pl Perl bindings for the original C implementation]
* [http://diotalevi.isa-geek.net/~josh/Jaro-Winkler/jjore's.pl Clean, perl reimplementation of the algorithm]