editDistance

Find the edit distance between the strings "Text analytics" and "Text analysis". The edit distance, by default, is the total number of grapheme insertions, deletions, and substitutions required to change one string to another.

str1 = "Text analytics";
str2 = "Text analysis";

Find the edit distance.

d = editDistance(str1,str2)

d = 
2

This means changing the first string to the second requires two edits. For example:

Find the edit distance between two tokenized documents. For tokenized document input, the edit distance, by default, is the total number of token insertions, deletions, and substitutions required to change one document to another.

str1 = "It's time for breakfast.";
document1 = tokenizedDocument(str1);

str2 = "It's now time to sleep.";
document2 = tokenizedDocument(str2);

Find the edit distance.

d = editDistance(document1,document2)

d = 
3

This means changing the first document to the second requires three edits. For example:

The editDistance function, by default, returns the lowest number of grapheme insertions, deletions, and substitutions required to change one string to another. To also include the swap action in the calculation, use the 'SwapCost' option.

First, find the edit distance between the strings "MATALB" and "MATLAB".

str1 = "MATALB";
str2 = "MATLAB";
d = editDistance(str1,str2)

d = 
2

One possible edit is:

The default value for the swap cost (the cost of swapping two adjacent graphemes) is Inf. This means that swaps do not count towards the edit distance. To include swaps, set the 'SwapCost' option to 1.

d = editDistance(str1,str2,'SwapCost',1)

d = 
1

This means there is one action. For example, swap the adjacent characters "A" and "L".

To compute the edit distance between two words and specify that the edits are case-insensitive, specify a custom substitute cost function.

First, compute the edit distance between the strings "MATLAB" and "MathWorks".

d = editDistance("MATLAB","MathWorks")

d = 
8

This means changing the first string to the second requires eight edits. For example:

Compute the edit distance and specify the custom substitution cost function caseInsensitiveSubstituteCost, listed at the end of the example. The custom function caseInsensitiveSubstituteCost returns 0 if the two inputs are the same or differ only by case and returns 1 otherwise.

d = editDistance("MATLAB","MathWorks",'SubstituteCost',@caseInsensitiveSubstituteCost)

d = 
6

This means the total cost for changing the first string to the second is 6. For example:

function cost = caseInsensitiveSubstituteCost(grapheme1,grapheme2)

if lower(grapheme1) == lower(grapheme2)
    cost = 0;
else
    cost = 1;
end

end