Jaro-Winkler distance

An algorithm to compute strings similarity / distance.

For an accurate description, see Wikipedia.

Usage

JaroWinkler.distance("qwerty", "qwertyu");
// output: 0.028571[...]

JaroWinkler.distance("foo", "");
// output: 1

JaroWinkler.similarity("one", "once");
// output: 0.933333[...]

JaroWinkler.similarity("One", "once", {caseSensitive: true});
// output: 0.722222[...]

API

The module has two method, one for similarity and one for distance

• They both accept the same input arguments
• They both return a value in the range [0, 1]
• They are complementary: similarity = 1 - distance (with perfect floating point precision)

JaroWinkler.similarity(string1, string2, inputConfig={})

Takes two strings as input, and an optionnal configuration, and compute their similarity.

It returns a value in value in the range [0, 1]. A value of 0 corresponds to no similarity, a value of 1 corresponds to complete similarity.

JaroWinkler.distance(string1, string2, inputConfig={})

Takes two strings as input, and an optionnal configuration, and compute their similarity.

It returns a value in value in the range [0, 1]. A value of 0 corresponds to no distance, a value of 1 corresponds to the maximum distance.

Configuration

• caseSensitive: boolean, indicates if the similarity should be case-sensitive (default: false)
• enableWinkler: boolean, indicates if the Winkler score modifier should be used (default: true)
• scalingFactor: float, set how much the score is adjusted to common prefixes (default: 0.25)
• prefixLength: integer, set the length of words prefix (default: 4)

This is the default config:

var config = {
  caseSensitive: false,
  enableWinkler: true,
  scalingFactor: 0.1,
  prefixLength: 4
};

Installation

The module can be installed from npm

npm install @ogus/jaro-winkler

It can also be installed by downloading the repo & including the jaro-winkler.js file in your project.

License

This project is licensed under the WTFPL - see LICENSE for more details