This is machine learning model and vectorizer. And the web application which is integrate with it
As this is text classification application used Naive Bayes classifier for train the model There are thre Naive Bayes classifiers I tested
- Multinomial Naive Bayes Classifier
- Gaussian Naive Bayes Calssifier
- Bernouli Naive Bayes Calssifier
EDA process has done with my own function, including:
- Lowercase the words
- Removing punctuation marks
- Remooving commonly used words which are not affect the performance of the model
Web application is created using Streamlit. Simple UI was created for ease of use of everyone.
- Download the repository
- Run this command in terminal (Make sure you have internet connection)
pip install -r requirements.txt
- Then in terminal you can simply use following command to use the application.
streamlit run app.py
- You can enter email text and predict for spam