HPR4026: Using NLP to get better answer options for language learning

Code snippets

Install the Levenshtein package.

>>> pip install levenshtein

Use the Levenshtein package to calculate the edit distance.

>>> import Levenshtein
>>> Levenshtein.distance("cat", "dog")
3
>>> Levenshtein.distance("cat", "can")
1

Install the TheFuzz package.

>>> pip install thefuzz

Use TheFuzz to calculate the similarity ratio.

>>> from thefuzz import fuzz
>>> fuzz.ratio("cat", "dog")
0
>>> fuzz.ratio("cat", "can")
67

Use TheFuzz to calculate the similarity ratio for each item in a list.

>>> from thefuzz import process
>>> options = ["dog", "can"]
>>> process.extract("cat", options, limit=2)
[('can', 67), ('dog', 0)]
>>> process.extract("cat", options, limit=1)
[('can', 67)]
>>> process.extract("cat", options, limit=2, scorer=fuzz.ratio)
[('can', 67), ('dog', 0)]

Other resources

• Ethan Nam's Understanding the Levenshtein Distance Equation for Beginners
• Levenstein package
• TheFuzz

Comments

More Information...