can use the cosine similarity between the two words, which can be done with first assigning vectors of the words using googles’ word2vec data, then calculating the cosine of the two vectors.
It gives you the semantic measure of similarity between the two words.
Do you mean something like WordNet based word similarity? Would a number between 0 & 1 be an appropriate response, where 1 is equal, 0 is no similarity?
can use the cosine similarity between the two words, which can be done with first assigning vectors of the words using googles’ word2vec data, then calculating the cosine of the two vectors.
It gives you the semantic measure of similarity between the two words.
Do you mean something like WordNet based word similarity? Would a number between 0 & 1 be an appropriate response, where 1 is equal, 0 is no similarity?
Semantic similarity that is (not some sort of distance).