NLP Vector Math

Project goal: make vector math valid for words using their embeddings

Natural Language Processing(NLP) process and analyze human language using computer programming.
Human language, like English consists of words and sentences.
NLP extracts information from these sentences.
In this project, I use words embeddings to transform words into numbers so that one can apply vector math.

NLP examples:

  • text classification : genre, autor, emotion

  • mistyping

  • page rank for search

  • text generation (text2image)

  • translation

  • text summary

  • dialog systems

Why embeddings?

Any computing machine operates using numbers.

Words embeddings allow us transform words into numbers.

As a result, we can use vector math on words:

Example.

King - Man + Women = Queen

(13, 5) - (12,0) + (-11,0) = (-10,5)