SMS/Email Spam Classifier

This Project is my final year project based on Machine Learning/Deep Learning/AI and is used to classify spam messages received through SMS or Email.

Technologies Used

• python
• pandas
• numpy
• flask
• scikit-learn
• nltk
• pickle

Why this technologies were used :

• python - Readability, Extensive, Libraries and Frameworks.
• pandas - Great way to read csv(comma separated values) and manipulate dataframes
• numpy - Used to perform array operations which are faster that python lists for operations
• flask - minimalistic python framework to host web application
• scikit-learn - python library for preprocessing and model training
• nltk - Python's language processing library for processing the data
• pickle - used to store the model, and preprocessor for the web application

How to install and run Project

Clone the repository in your system

git clone https://github.com/Simpl1fy/Spam-Classifier-Project.git

Open the directory in VSCode terminal(cmd) or CommandPrompt(cmd) and

pip install virtualenv
virtualenv venv
venv\Scripts\activate.bat

After enabling the Virtual Enviornment

pip install -r requirements.txt

This may take some time based on your network speed and speed of your system. After installation you may run the server

python app.py

After running the server you can copy the link to a new tab in your web browser

localhost:5000

Problems faced while making the project

• Using Javascript for creating a single form that can take both for SMS and Email.
• Printing output as a flask in the html file
• Long wait times while training the data of Email Spam Classifier, due to huge amount of data
• Taking input from the form and creating data to predict using the model

Improvements that can be done

• Can be created into package
• The model can be self trained where the data will be stored in the database and new data will be stored in the database and trained every data for new model with better accuracies
• Implementing the project into AWS with CICD pipeline