“King – Man + Woman = ?”
A word vector is a continuous word representation wherein the vector’s weight is distributed across various elements. So, instead of having a one-to-one mapping between a vector element and a word, the representation of a word is spread across all of the elements in the vector, and each element in the vector contributes to the definition of many words. These vectors somehow represent the “meaning” of a word.
These vectors are good in answering analogies like man:woman as uncle:____ (aunt) by using a simple vector offset method based on cosine distance. With this, we can answer “King – Man + Woman = ?” question and arrive at the result “Queen”. This shows that using a word offset technique where simple algebraic operations are performed on the word vectors results in a vector that is closest to the vector representation of the word “Queen”.
Vectors for King, Man, Queen, & Woman:
The result of the vector composition King – Man + Woman = ?
Word vectors with such semantic relationships could be used to improve many existing Natural Language Processing applications, such as machine translation, information retrieval and question answering systems, and may enable other future applications yet to be invented.