TextClassification-Keras

Why we included this project

This repository collects the classic deep learning architectures for text classification, all implemented in Keras and laid out so you can compare them directly. FastText, TextCNN, TextRNN, bidirectional and attention-augmented RNN variants, HAN, and RCNN each live in their own directory with a small runnable application, and the README links every implementation back to the paper that introduced it. Because the models share the same framework and follow the same structure, the trade-offs between them are easy to see: where one encodes text with convolutions, another leans on recurrent layers or attention. The code is readable enough to adapt to your own dataset, which makes this a practical starting point for students learning NLP or teams prototyping a classifier. It is a study aid, not a production library.