Discriminate edit distance

Question

The levenshtein edit distance cares only about how many edits are done and not on what exactly they are, so the following two pairs will have the same edit distance.

("A P Moller - Maersk A", "A.P. Moller - Maersk A/S Class A")
("A P Moller - Maersk A", "A.P. Moller - Maersk A/S Class B")

Are there any algorithms or libraries that can distinguish between these two pairs?

Answer 1

You can use cosine similarity to find the similarity between to text and it produce different similarity between these two text

import math
import re
from collections import Counter

WORD = re.compile(r"\w+")


def get_cosine(vec1, vec2):
    intersection = set(vec1.keys()) & set(vec2.keys())
    numerator = sum([vec1[x] * vec2[x] for x in intersection])
    sum1 = sum([vec1[x] ** 2 for x in list(vec1.keys())])
    sum2 = sum([vec2[x] ** 2 for x in list(vec2.keys())])
    denominator = math.sqrt(sum1) * math.sqrt(sum2)
    if not denominator:
        return 0.0
    else:
        return float(numerator) / denominator


def text_to_vector(text):
    words = WORD.findall(text)
    return Counter(words)

x =("A P Moller - Maersk A", "A.P. Moller - Maersk A/S Class A")
y =("A P Moller - Maersk A", "A.P. Moller - Maersk A/S Class B")

cosine = get_cosine(text_to_vector(x[0]), text_to_vector(x[1]))

print("Cosine1:", cosine)

cosine1 = get_cosine(text_to_vector(y[0]), text_to_vector(y[1]))

print("Cosine2:", cosine1)

Output:

Cosine1: 0.9091372900969896
Cosine2: 0.8366600265340756