Python for NLP: Movie Sentiment Analysis using Deep Learning in Keras
This is the 17th article in my series of articles on Python for NLP. In the last article, we started our discussion about deep learning for natural language processing.
The previous article was focused primarily towards word embeddings, where we saw how the word embeddings can be used to convert text to a corresponding dense vector, which can be subsequently used as input to any deep learning model. We perform basic classification task using word embeddings. We used custom dataset that contained 16 imaginary reviews about movies. Furthermore, the classification algorithms were trained and tested on same data. Finally, we only used a densely connected neural network to test our algorithm.
In this article, we will build upon the concepts that we studied in the previous article and will see classification in more detail using a real-world dataset. We will use three different types of deep neural networks: Densely connected neural network (Basic Neural Network), Convolutional Neural Network (CNN) and Long Short Term Memory Network (LSTM), which is a variant of Recurrent Neural Networks. Furthermore, we will see how to evaluate deep learning model on a totally unseen