In this example, Kedro was used to create a workflow to achieve the following:
- Pre-process the Sentiment140 dataset containing 1.6 million tweets
- Train a Logistic Regression model for sentiment prediction based on the text of the tweets
- Evaluate the trained model and generate a classification report
This project was created as part of the PyCon India 2020 Devsprint.
Special thanks to Lais Carvalho for her guidance through the devsprint.