There's a method [most_similar()][1] that will report the words of the closest vectors, by cosine-similarity in the model's coordinates, to a given word.
確定! 回上一頁