This is a code repository collection for the NLP Machine Learning Spam Classifier end-to-end project. The article for this code will be attached here as soon as it has been published. The code is still incomplete as it's still in the process of cleaning up, so keep an eye on this repository.

Aim

This project aims to develop a spam classifier model from the available email data and simulate how to deploy and maintain them.

Procedure

We would follow the standard data science project for this project, including:

1. Data Exploration
2. Data Preprocessing
3. Model Development
4. Model Evaluation
5. Model Deployment
6. Model Maintenance

The overall project structure would look like the following:

Tech Stack

For this project, we would use various open-source tech stacks that are appropriate for each step, including:

1. Python
2. Jupyter Notebook
3. VS Code
4. Docker
5. Airflow
6. FastAPI

Additionally, we would use related Python packages common in data science projects. Some notable packages include:

1. MLFlow
2. Evidently
3. NLTK
4. Wordcloud

Dataset

You can find the dataset in the storage folder (https://dagshub.com/cornelliusyudhawijaya/End_to_End_Spam_Classifier/src/master/storage/spam_emails.csv

Screenshot

Some of the screenshots of this project activity.

1. Email Statistics

2. Visualization of the Email Statistics

3. Word cleanings and word cloud visualization

4. Model development and evaluation experiment

5. Model Evaluation various metrics change in each iteration

6. Model Front-end Streamlit simple dashboard

7. Building the model API and serving in the backend

8. Using Docker Compose to containerize the frontend and backend

9. Tracking the model experiments with MLFlow