Summary: Text Classification with SpaCy

We have finished discussing a very hot NLP topic—text classification.

Key points

• We learned about text classification concepts such as binary classification, multilabel classification, and multiclass classification.

• We learned how to train TextCategorizer, spaCy's text classifier component.

• We learned how to transform your data into spaCy training format and then train the TextCategorizer component with this data.

• We learned how to combine spaCy code and Keras code.

• We learned the basics of neural networks, including some handy layers such as the dense layer, dropout layer, embedding layer, and recurrent layers.

• We learned how to tokenize and preprocess the data with Keras' Tokenizer.

• We went through neural network design with tf.keras code and learned how to design and evaluate a statistical experiment with LSTM.