Let’s say we have 300 genres, that would mean each column of the table would be a 300×1 dimension vector representing the nuance of that word. Vectorizing this nuance is called word-embedding.
By making the machine learn this nuance(the table above), now we can make the machine understand human language.
