Emotion Recognition App

Emotion Recognition is an app made with Streamlit in Python, and using NLP, which purpose is to find emotions in text, by using machine learning and deep learning algorithms on two datasets, and compare the different techniques with metrics.

https://share.streamlit.io/atelders/emotion_recognition/main/src/main.py

Click on Classification, Neural Network and fastText to test with a custom sentence.

Datasets

Two datasets were used:

From Kaggle : Emotion_final.csv From data.world : text_emotion.csv

Source Code

Source code of the app is in src/main.py

Libraries

streamlit, pandas, matplotlib.pyplot, seaborn, fasttext, io, pickle, pathlib, sklearn, tensorflow.keras

File Structure

├── assets
│   ├── frequency_kaggle.png
│   └── frequency_world.png
├── data
│   ├── models
│   │   ├── dt_kaggle_count.sav
│   │   ├── dt_kaggle_tfidf.sav
│   │   ├── dt_world_count.sav
│   │   ├── dt_world_tfidf.sav
│   │   ├── lr_kaggle_count.sav
│   │   ├── lr_kaggle_tfidf.sav
│   │   ├── lr_world_count.sav
│   │   ├── lr_world_tfidf.sav
│   │   ├── svm_kaggle_count.sav
│   │   ├── svm_kaggle_tfidf.sav
│   │   ├── svm_world_count.sav
│   │   └── svm_world_tfidf.sav
│   └── raw
│       ├── Emotion_final.csv
│       └── text_emotion.csv
├── environment.yml
├── fasttext_data.world.bin
├── fasttext_data.world binary.bin
├── fasttext_Kaggle.bin
├── LICENSE
├── notebooks
│   ├── 20210517-at-cs-jd-initial-data-exploration.ipynb
│   ├── neurons.ipynb
│   ├── Random_forest.ipynb
│   └── words_frequency.ipynb
├── README.md
├── src
│   ├── __init__.py
│   ├── main.py
│   └── words_frequency.py
├── test.txt
└── train.txt