PredEmotions

A neural network dedicated to sentiment classification based on the underlying emotions of a sentence.

Usage

1. Install the requirements using pip install -r requirements.txt
2. Getting a model
   • Using a pretrained model: Move the model to trained/model.pt or use the existing model in the repository
   • Training a model:
     1. The dataset should placed in data.
     2. Run python3 setupy.py, it will explain the rest:
3. Running the classifier: Running the command python3 setup.py and hitting enter will start the inference loop.

Using in Your Program

This is more of a proof of concept so there is no package to install this and directly integrate it. To use it the source code can be included in your program and once the dataset and model are in the proper location you can include the prediction pipeline with from {Insert Source Location}.model.Predictor import Predictor and run Predictor().predict(string). Feel free to adapt and modify the code to your needs.

Datasets Being Used

• Text to Emotion Classification