R - The type of similarity score unit used by this EditDistance.EditDistance<R>CosineDistance, HammingDistance, JaccardDistance, JaccardSimilarity, JaroWinklerDistance, LevenshteinDetailedDistance, LevenshteinDistance, LongestCommonSubsequence, LongestCommonSubsequenceDistancepublic interface SimilarityScore<R>
A string similarity score is intended to have some of the properties of a metric, yet allowing for exceptions, namely the Jaro-Winkler similarity score.
We Define a SimilarityScore to be a function d: [X * X] -> [0, INFINITY) with the
following properties:
d(x,y) >= 0, non-negativity or separation axiomd(x,y) == d(y,x), symmetry.Notice, these are two of the properties that contribute to d being a metric.
Further, this intended to be BiFunction<CharSequence, CharSequence, R>.
The apply method
accepts a pair of CharSequence parameters
and returns an R type similarity score. We have omitted the explicit
statement of extending BiFunction due to it only being implemented in Java 1.8, and we
wish to maintain Java 1.7 compatibility.
| Modifier and Type | Method | Description |
|---|---|---|
R |
apply(CharSequence left,
CharSequence right) |
Compares two CharSequences.
|
R apply(CharSequence left, CharSequence right)
left - the first CharSequenceright - the second CharSequenceCopyright © 2014–2018 The Apache Software Foundation. All rights reserved.