The aim of this project is to check if the model trained on generic tweets works well in determining the public sentiment about the political parties provide any insight about the outcome of the 2019 Canadian Elections.
1. Import required libraries and create a dataframe of the CSV file
2. Comprehensive text preprocessing is implemented on the generic tweets
3. Enough text preprocessing is implemented such that we can classify the tweets for each party, after classifying we clean the tweets completely
4. Various text vectorizing techniques coupled with various models are implemented and the model performance evaluated
5. The best model is then implemented on Canadian elections tweets, model performance and the factors affecting the model performance are evaluated
6. Feature engineering is implemented to better classify the tweets
7. Various insights are derived from the data