After exploring the dataset, it is time to build the RNN model to predict the sentiment of movie reviews from data recorded in the dataset. Before proceeding, it is necessary to prepare the data appropriately. This is a question of solving the problem of how words can be coded to process textual information automatically, in a similar way to what is done on images, as we saw in Chapter 5, Fashion Article Recognition using Convolutional Neural Networks.
One possible solution is represented by word embeddings, a set of tools, language models, and machine learning techniques that allow the representation of words and phrases through the use of real-component vectors. Keras provides a special layer to perform ...
