【发布时间】：2013-10-01 18:39:55
【问题描述】：
如何在 C# 中实现 Jaro-Winkler 距离字符串比较算法？
【问题讨论】：
标签： string comparison distance jaro-winkler
【发布时间】：2013-10-01 18:39:55
【问题描述】：
public static class JaroWinklerDistance
{
/* The Winkler modification will not be applied unless the
* percent match was at or above the mWeightThreshold percent
* without the modification.
* Winkler's paper used a default value of 0.7
*/
private static readonly double mWeightThreshold = 0.7;
/* Size of the prefix to be concidered by the Winkler modification.
* Winkler's paper used a default value of 4
*/
private static readonly int mNumChars = 4;
/// <summary>
/// Returns the Jaro-Winkler distance between the specified
/// strings. The distance is symmetric and will fall in the
/// range 0 (perfect match) to 1 (no match).
/// </summary>
/// <param name="aString1">First String</param>
/// <param name="aString2">Second String</param>
/// <returns></returns>
public static double distance(string aString1, string aString2) {
return 1.0 - proximity(aString1,aString2);
}
/// <summary>
/// Returns the Jaro-Winkler distance between the specified
/// strings. The distance is symmetric and will fall in the
/// range 0 (no match) to 1 (perfect match).
/// </summary>
/// <param name="aString1">First String</param>
/// <param name="aString2">Second String</param>
/// <returns></returns>
public static double proximity(string aString1, string aString2)
{
int lLen1 = aString1.Length;
int lLen2 = aString2.Length;
if (lLen1 == 0)
return lLen2 == 0 ? 1.0 : 0.0;
int lSearchRange = Math.Max(0,Math.Max(lLen1,lLen2)/2 - 1);
// default initialized to false
bool[] lMatched1 = new bool[lLen1];
bool[] lMatched2 = new bool[lLen2];
int lNumCommon = 0;
for (int i = 0; i < lLen1; ++i) {
int lStart = Math.Max(0,i-lSearchRange);
int lEnd = Math.Min(i+lSearchRange+1,lLen2);
for (int j = lStart; j < lEnd; ++j) {
if (lMatched2[j]) continue;
if (aString1[i] != aString2[j])
continue;
lMatched1[i] = true;
lMatched2[j] = true;
++lNumCommon;
break;
}
}
if (lNumCommon == 0) return 0.0;
int lNumHalfTransposed = 0;
int k = 0;
for (int i = 0; i < lLen1; ++i) {
if (!lMatched1[i]) continue;
while (!lMatched2[k]) ++k;
if (aString1[i] != aString2[k])
++lNumHalfTransposed;
++k;
}
// System.Diagnostics.Debug.WriteLine("numHalfTransposed=" + numHalfTransposed);
int lNumTransposed = lNumHalfTransposed/2;
// System.Diagnostics.Debug.WriteLine("numCommon=" + numCommon + " numTransposed=" + numTransposed);
double lNumCommonD = lNumCommon;
double lWeight = (lNumCommonD/lLen1
+ lNumCommonD/lLen2
+ (lNumCommon - lNumTransposed)/lNumCommonD)/3.0;
if (lWeight <= mWeightThreshold) return lWeight;
int lMax = Math.Min(mNumChars,Math.Min(aString1.Length,aString2.Length));
int lPos = 0;
while (lPos < lMax && aString1[lPos] == aString2[lPos])
++lPos;
if (lPos == 0) return lWeight;
return lWeight + 0.1 * lPos * (1.0 - lWeight);
}
}
【讨论】：
-
戴夫，这段代码实际上支持两者。调用proximity(str1, str2) 返回0-1，其中1 是完美匹配。调用 distance(str1, str2) 返回 0-1，其中 0 为完美匹配。
-
我对您的实现做了一些（有用的）修改：gist.github.com/ronnieoverby/2aa19724199df4ec8af6 最有用的是指定字符比较器的能力。
-
这匹配 mdq Microsoft.MasterDataServices.DataQuality 程序集 Jaro-Winkler 的分数比任何可用的也适用 Jaro-Winkler 的 Nuget 包都要好。
-
我也冒昧地将它移植到 TSQL，如果有人感兴趣，您可以在 stackoverflow.com/a/34879177/1856391 找到代码 :)
-
是的。一些 Nuget 包存在错误，这些错误会导致错误的 Jaro Winkler “相似性”并存在无限循环问题。 fuzzystring.codeplex.com/workitem/list/basic
你可以去Lucene.Net看看，
它实现了 Jaro-Winkler 距离算法，
而且它的分数和哪个leebickmtu post不一样，
可以参考
网址如下：
http://lucenenet.apache.org/docs/3.0.3/db/d12/_jaro_winkler_distance_8cs_source.html
【讨论】：
-
我在 Lucene.Net 上尝试了代码，它给出的结果与我发布的解决方案相同。它与我的代码的接近（）调用相匹配。 distance() 调用只是反转间隔。
-
你是如何实现这段代码的，任何例子都会对我非常有帮助。
您可以使用下面的代码，它对所有类型的字符串都非常有效。得到结果后，您需要乘以 100 以获得相似度百分比。希望能解决你的问题。
public class JaroWinkler
{
private const double defaultMismatchScore = 0.0;
private const double defaultMatchScore = 1.0;
/// <summary>
/// Gets the similarity between two strings by using the Jaro-Winkler algorithm.
/// A value of 1 means perfect match. A value of zero represents an absolute no match
/// </summary>
/// <param name="_firstWord"></param>
/// <param name="_secondWord"></param>
/// <returns>a value between 0-1 of the similarity</returns>
///
public static double RateSimilarity(string _firstWord, string _secondWord)
{
// Converting to lower case is not part of the original Jaro-Winkler implementation
// But we don't really care about case sensitivity in DIAMOND and wouldn't decrease security names similarity rate just because
// of Case sensitivity
_firstWord = _firstWord.ToLower();
_secondWord = _secondWord.ToLower();
if ((_firstWord != null) && (_secondWord != null))
{
if (_firstWord == _secondWord)
//return (SqlDouble)defaultMatchScore;
return defaultMatchScore;
else
{
// Get half the length of the string rounded up - (this is the distance used for acceptable transpositions)
int halfLength = Math.Min(_firstWord.Length, _secondWord.Length) / 2 + 1;
// Get common characters
StringBuilder common1 = GetCommonCharacters(_firstWord, _secondWord, halfLength);
int commonMatches = common1.Length;
// Check for zero in common
if (commonMatches == 0)
//return (SqlDouble)defaultMismatchScore;
return defaultMismatchScore;
StringBuilder common2 = GetCommonCharacters(_secondWord, _firstWord, halfLength);
// Check for same length common strings returning 0 if is not the same
if (commonMatches != common2.Length)
//return (SqlDouble)defaultMismatchScore;
return defaultMismatchScore;
// Get the number of transpositions
int transpositions = 0;
for (int i = 0; i < commonMatches; i++)
{
if (common1[i] != common2[i])
transpositions++;
}
int j = 0;
j += 1;
// Calculate Jaro metric
transpositions /= 2;
double jaroMetric = commonMatches / (3.0 * _firstWord.Length) + commonMatches / (3.0 * _secondWord.Length) + (commonMatches - transpositions) / (3.0 * commonMatches);
//return (SqlDouble)jaroMetric;
return jaroMetric;
}
}
//return (SqlDouble)defaultMismatchScore;
return defaultMismatchScore;
}
/// <summary>
/// Returns a string buffer of characters from string1 within string2 if they are of a given
/// distance seperation from the position in string1.
/// </summary>
/// <param name="firstWord">string one</param>
/// <param name="secondWord">string two</param>
/// <param name="separationDistance">separation distance</param>
/// <returns>A string buffer of characters from string1 within string2 if they are of a given
/// distance seperation from the position in string1</returns>
private static StringBuilder GetCommonCharacters(string firstWord, string secondWord, int separationDistance)
{
if ((firstWord != null) && (secondWord != null))
{
StringBuilder returnCommons = new StringBuilder(20);
StringBuilder copy = new StringBuilder(secondWord);
int firstWordLength = firstWord.Length;
int secondWordLength = secondWord.Length;
for (int i = 0; i < firstWordLength; i++)
{
char character = firstWord[i];
bool found = false;
for (int j = Math.Max(0, i - separationDistance); !found && j < Math.Min(i + separationDistance, secondWordLength); j++)
{
if (copy[j] == character)
{
found = true;
returnCommons.Append(character);
copy[j] = '#';
}
}
}
return returnCommons;
}
return null;
}
}
【讨论】：
使用下面的类来使用 jaro winkler。 我已经自定义了算法 jaro 和 jaro-winkler。
访问Github 获取 DLL。
using System;
using System.Linq;
namespace Search
{
public static class EditDistance
{
private struct JaroMetrics
{
public int Matches;
public int Transpositions;
}
private static EditDistance.JaroMetrics Matches(string s1, string s2)
{
string text;
string text2;
if (s1.Length > s2.Length)
{
text = s1;
text2 = s2;
}
else
{
text = s2;
text2 = s1;
}
int num = Math.Max(text.Length / 2 - 1, 0);
int[] array = new int[text2.Length];
int i;
for (i = 0; i < array.Length; i++)
{
array[i] = -1;
}
bool[] array2 = new bool[text.Length];
int num2 = 0;
for (int j = 0; j < text2.Length; j++)
{
char c = text2[j];
int k = Math.Max(j - num, 0);
int num3 = Math.Min(j + num + 1, text.Length);
while (k < num3)
{
if (!array2[k] && c == text[k])
{
array[j] = k;
array2[k] = true;
num2++;
break;
}
k++;
}
}
char[] array3 = new char[num2];
char[] ms2 = new char[num2];
i = 0;
int num4 = 0;
while (i < text2.Length)
{
if (array[i] != -1)
{
array3[num4] = text2[i];
num4++;
}
i++;
}
i = 0;
num4 = 0;
while (i < text.Length)
{
if (array2[i])
{
ms2[num4] = text[i];
num4++;
}
i++;
}
int num5 = array3.Where((char t, int mi) => t != ms2[mi]).Count<char>();
EditDistance.JaroMetrics result;
result.Matches = num2;
result.Transpositions = num5 / 2;
return result;
}
public static float JaroWinkler(this string s1, string s2, float prefixScale, float boostThreshold)
{
prefixScale = ((prefixScale > 0.25f) ? 0.25f : prefixScale);
prefixScale = ((prefixScale < 0f) ? 0f : prefixScale);
float num = s1.Jaro(s2);
int num2 = 0;
for (int i = 0; i < Math.Min(s1.Length, s2.Length); i++)
{
if (s1[i] != s2[i])
{
break;
}
num2++;
}
return (num < boostThreshold) ? num : (num + prefixScale * (float)num2 * (1f - num));
}
public static float JaroWinkler(this string s1, string s2, float prefixScale)
{
return s1.JaroWinkler(s2, prefixScale, 0.7f);
}
public static float JaroWinkler(this string s1, string s2)
{
return s1.JaroWinkler(s2, 0.1f, 0.7f);
}
public static float Jaro(this string s1, string s2)
{
EditDistance.JaroMetrics jaroMetrics = EditDistance.Matches(s1, s2);
float num = (float)jaroMetrics.Matches;
int transpositions = jaroMetrics.Transpositions;
float result;
if (num == 0f)
{
result = 0f;
}
else
{
float num2 = (num / (float)s1.Length + num / (float)s2.Length + (num - (float)transpositions) / num) / 3f;
result = num2;
}
return result;
}
public static int LevenshteinDistance(this string source, string target)
{
int result;
if (string.IsNullOrEmpty(source))
{
if (string.IsNullOrEmpty(target))
{
result = 0;
}
else
{
result = target.Length;
}
}
else if (string.IsNullOrEmpty(target))
{
result = source.Length;
}
else
{
if (source.Length > target.Length)
{
string text = target;
target = source;
source = text;
}
int length = target.Length;
int length2 = source.Length;
int[,] array = new int[2, length + 1];
for (int i = 1; i <= length; i++)
{
array[0, i] = i;
}
int num = 0;
for (int j = 1; j <= length2; j++)
{
num = (j & 1);
array[num, 0] = j;
int num2 = num ^ 1;
for (int i = 1; i <= length; i++)
{
int num3 = (target[i - 1] == source[j - 1]) ? 0 : 1;
array[num, i] = Math.Min(Math.Min(array[num2, i] + 1, array[num, i - 1] + 1), array[num2, i - 1] + num3);
}
}
result = array[num, length];
}
return result;
}
}
}
【讨论】：