How can I implement word comparison taking into account word forms in Golang?

Question

go

How can words be compared taking into account word forms? Are there ready-made algorithms that allow you to get an abstract function that will work like thiscompareWords(first string, second string) bool

compareWords("автомобиль", "автомобильный") == true
compareWords("автомобиль", "aвиационный") == false

Levenshtein distancedid not fit, since this algorithm, for example, considers the words инициацияand to be similar ИИ. Too many false positives.

Answer 1

As already answered, you need a stemmer to normalize words.
Concrete implementation in Go:
https://github.com/blevesearch/snowballstem

Answer 2

Keywords for a more or less correct solution to the problem: stemming, lemmatization, and NLP in general