Sentiment Analysis in Reviews using Natural Language Processing and Word Embeddings and interprets each word as a vector in an N-dimensional space. The Network can predict whether any review is positive or negative. Each word is mapped on a spherical figure
Using the IMDB dataset from stanford that can be found here : https://ai.stanford.edu/~amaas/data/sentiment/
Dataset is made up of 50K movie reviews from imdb.
It uses Word Embeddings i.e. it interprets each word as a vector in a N-dimensional space and Similar words point to similar directions. It is a Binary Neural Network made using Tensorflow.