Finding emotions from text with 88.9% accuracy in a corpus of almost 22 thousand unique sentences. Classifying 6 different emotion from text by using a Sequentially trained Keras model having Bi-Directional LSTM layer.
Run each section in one of the following files:
Small_Dataset.ipynb
Medium_dataset.ipynb
Huge_Dataset.ipynb
Results
Accuracy and Accuracy Graphs
Model Summary and Model Loss
Tools / Libraries
1: Tensorflow
2: Sklearn
3: NLTK
4: Seaborn
5: Pandas
6: Matplotlib
7: Numpy