[
https://issues.apache.org/jira/browse/TEXT-126?focusedWorklogId=206850&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-206850
]
ASF GitHub Bot logged work on TEXT-126:
---------------------------------------
Author: ASF GitHub Bot
Created on: 03/Mar/19 05:23
Start Date: 03/Mar/19 05:23
Worklog Time Spent: 10m
Work Description: kinow commented on pull request #103: TEXT-126: Adding
Sorensen-Dice similarity algoritham
URL: https://github.com/apache/commons-text/pull/103#discussion_r261852025
##########
File path:
src/main/java/org/apache/commons/text/similarity/SorensenDicesSimilarity.java
##########
@@ -0,0 +1,118 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.commons.text.similarity;
+
+import java.util.HashSet;
+import java.util.Set;
+
+/**
+ * A similarity algorithm indicating the percentage of matched characters
+ * between two character sequences.
+ *
+ * <p>
+ * The Sørensen–Dice coefficient is a statistic used for comparing the
+ * similarity of two samples. It was independently developed by the botanists
+ * Thorvald Sørensen and Lee Raymond Dice, who published in 1948 and 1945
+ * respectively. The index is known by several other names, especially
+ * Sørensen–Dice index, Sørensen index and Dice's coefficient. Other variations
+ * include the "similarity coefficient" or "index", such as Dice similarity
+ * coefficient (DSC).
+ * </p>
+ *
+ * <p>
+ * This implementation is based on the Sørensen–Dice similarity algorithm from
+ * <a href=
+ * "http://en.wikipedia.org/wiki/S%C3%B8rensen%E2%80%93Dice_coefficient";>
+ * http://en.wikipedia.org/wiki/S%C3%B8rensen%E2%80%93Dice_coefficient</a>.
+ *
+ *
+ * </p>
+ *
+ * @since 1.7
+ */
+public class SorensenDicesSimilarity implements SimilarityScore<Double> {
+
+ /**
+ *
+ * <pre>
+ * similarity.apply("", "") = 1.0
+ * similarity.apply("foo", "foo") = 1.0
+ * similarity.apply("foo", "foo ") = 0.8
+ * similarity.apply("foo", "foo ") = 0.66
+ * similarity.apply("foo", " foo ") = 0.66
+ * similarity.apply("foo", " foo") = 0.66
+ * similarity.apply("", "a") = 0.0
+ * similarity.apply("aaapppp", "") = 0.0
+ * similarity.apply("frog", "fog") = 0.4
+ * similarity.apply("fly", "ant") = 0.0
+ * similarity.apply("elephant", "hippo") = 0.0
+ * similarity.apply("hippo", "elephant") = 0.0
+ * similarity.apply("hippo", "zzzzzzzz") = 0.0
+ * similarity.apply("hello", "hallo") = 0.5
+ * similarity.apply("ABC Corporation", "ABC Corp") = 0.7
+ * similarity.apply("D N H Enterprises Inc", "D & H Enterprises,
Inc.") = 0.74
+ * similarity.apply("My Gym Children's Fitness Center", "My Gym. Childrens
Fitness") = 0.81
+ * similarity.apply("PENNSYLVANIA", "PENNCISYLVNIA") = 0.69
+ * </pre>
+ *
+ * @param left the first CharSequence, must not be null
+ * @param right the second CharSequence, must not be null
+ * @return result similarity
+ * @throws IllegalArgumentException if either CharSequence input is {@code
null}
+ */
+ @Override
+ public Double apply(final CharSequence left, final CharSequence right) {
+
+ if (left == null || right == null) {
+ throw new IllegalArgumentException("CharSequences must not be
null");
+ }
+
+ if (left.equals(right)) {
+ return 1d;
+ }
+
+ if (left.length() == 0 || right.length() == 0) {
+ return 0d;
+ }
+
+ Set<String> nLeft = createBigrams(left);
+ Set<String> nRight = createBigrams(right);
+
+ final int total = nLeft.size() + nRight.size();
+ nLeft.retainAll(nRight);
Review comment:
It may not be necessary to call `retainAll` I think? We want to
`intersection` being the number of elements common to both
vectors/sets/sides/etc.
Indeed, calling `retainAll` does the trick here. But I believe
`createBigrams` is returning `HashSet`, which is backed - as far as I can tell
- by a `HashMap`, where get/put should be quite good, but not so much delete
operations (I remember looking at one implementation some time ago, and it had
so many checks in place :eyes: ).
So _I think_ we can achieve the same by just filtering the collections and
counting the total number of elements common to both sets?
e.g.
```diff
diff --git
a/src/main/java/org/apache/commons/text/similarity/SorensenDicesSimilarity.java
b/src/main/java/org/apache/commons/text/similarity/SorensenDicesSimilarity.java
index d4df1d7..c79023a 100644
---
a/src/main/java/org/apache/commons/text/similarity/SorensenDicesSimilarity.java
+++
b/src/main/java/org/apache/commons/text/similarity/SorensenDicesSimilarity.java
@@ -18,6 +18,7 @@ package org.apache.commons.text.similarity;
import java.util.HashSet;
import java.util.Set;
+import java.util.stream.Collectors;
/**
* A similarity algorithm indicating the percentage of matched characters
@@ -93,8 +94,9 @@ public class SorensenDicesSimilarity implements
SimilarityScore<Double> {
Set<String> nRight = createBigrams(right);
final int total = nLeft.size() + nRight.size();
- nLeft.retainAll(nRight);
- final int intersection = nLeft.size();
+ final long intersection = nLeft.stream()
+ .filter(nRight::contains)
+ .collect(Collectors.counting());
return (2.0d * intersection) / total;
}
```
I believe this way we take advantage of Java 8 syntax, and avoid creating a
new data structure, or performing multiple checks. The tests worked, but being
quite late Sunday for me, almost time to call it a day, and my last comment in
this pull request, I may be saying something totally silly. So no hard feelings
if anyone corrects me :+1:
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 206850)
Time Spent: 6.5h (was: 6h 20m)
> Dice's Coefficient Algorithm in String similarity
> -------------------------------------------------
>
> Key: TEXT-126
> URL: https://issues.apache.org/jira/browse/TEXT-126
> Project: Commons Text
> Issue Type: Improvement
> Reporter: Vicky Chawda
> Priority: Major
> Time Spent: 6.5h
> Remaining Estimate: 0h
>
> I'd like to propose an extension to the algorithms for string similarity in
> *commons-text/src/main/java/org/apache/commons/text/similarity/*
> Dice's Coefficient Algorithm can be helpful for many who are looking for
> ranking similarities in strings.
> *Inspired from* - [http://www.catalysoft.com/articles/StrikeAMatch.html]
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
