karan1pandya / sentipy

Sentiment analyzer for short text

Sentipy

This package is a basic sentiment analyzer

About sentipy

Sentipy attempts to classify a text into positive, negative or neutral sentiment. It uses Spacy's textcat with ensemble architecture in the back-end. The ultimate objective of this package is to classify the sentiments as accurately as possible

While the sentiment analysis at the core is absolutely basic, the current focus is to understand the features the model is learning. sentipy leverages on lime to get the features learnt and uses streamlit to crate a simple webapp that helps with the visualization

How to install sentipy

1. git clone the repo in to your local system
2. run setup.py install

How to run the feature visualizer app

On your terminal run sentipy streamlit

What next?

1. Use transformers and take the self supervised learning approach for classification
2. Include visualizations for pre-processed text
3. Make pre-processing options available on web app
4. Better visualization

test changes made by me