search

colonialjelly / name-matching

1 stars 1 forks source link

Levenshtein edit distance #3

Open DallanQ opened 3 years ago

DallanQ commented 3 years ago

Another baseline that could be considered is a normalized levenshtein edit distance: 1 - (edit distance / len(longest string)

colonialjelly commented 3 years ago

That's a good idea. Here's a cool library I've used in the past for string similarity stuff https://jellyfish.readthedocs.io/en/latest it has implementations of a lot of the standard string matching algorithms.

DallanQ commented 3 years ago

I like that it has many different algorithms we can try