Word2vec explained: deriving Mikolov et al.'s negative-sampling word-embedding method

I recently received a task to study word2vec. I felt that the layers of information on the Network were not neat, And I always felt that the explanation was not so complete. Maybe even the author himself does not know why his model is easy to use. The negtive sampling mentioned in this paper has given me a lot of confusion. The following Connell article has inspired me a lot and clearly provided a method for understanding negtive Sampling:

Word2vec explained: deriving Mikolov et al.'s negative-sampling word-embedding method

The author of this article answered a very humorous question: Why is word2vec so helpful? The answer is: We don't know, but he can use it. The answer is the same as that of the child I handed in.