Introduction

In this section you will be building a bidirectional LSTM for text classification. The bidirectional LSTM you build in this section will be trained on a text corpus of movie reviews, separated into positive and negative reviews.

A. Classifying text

One of the most important aspects of NLP is the task of classifying texts into different categories. Applications of text classification include spam filtering, classifying user product reviews, and automatically flagging inappropriate or harmful social media posts.

Text classification in the form of spam filtering. In this example, you see an email get classified as spam.
Text classification in the form of spam filtering. In this example, you see an email get classified as spam.

When dealing with text data it is almost always better to use an LSTM over a regular feed-forward neural network. This is because the LSTM is built specifically for sequential data like text data.

Therefore we'll use a variation known as the bidirectional LSTM for our text classification in the upcoming chapters.

Create a free account to view this lesson.

Continue your learning journey with a 14-day free trial.

By signing up, you agree to Educative's Terms of Service and Privacy Policy