Custom Text Classification on Android using TensorFlow Lite
A lot of social media platforms have been using AI these days to classify vulgar and offensive posts and automatically take them down. I thought why not try doing something similar; and so, I’ve come up with this end-to-end tutorial that will help you build your own corpus for training a text classification model, and later export and deploy it on an Android app for you to use. All this, absolutely on a custom dataset of your choice.
So, are you excited to build your own text classifier app? If yes, let’s begin the show.
Now, before we begin, let me tell you that we’ll be doing all the model’s hyperparameter configuration and training on Google Colab. To build the Android app, we’ll need to have Android Studio. If you haven’t installed it yet, find it here.
Step 1: Build a corpus
I’m building a text classifier that