Email/SMS Spam Classifier

Table of Contents

• Introduction
• Screenshots
• Deployed Website
• Features
• Technologies Used
• Installation
• Contributions
• License

Introduction

The main objective behind this application is to classify SMS/Email as spam or not spam using machine learning. With an ever-growing demand to communicate in professional settings such as institutes and workplaces, especially during times when remote communication became essential, the risk of encountering spam has significantly increased.

Unfortunately, this environment also provides fertile ground for spammers who exploit it to execute fraudulent activities. To combat this, our application utilizes Python, a powerful tool for building machine learning models, to help distinguish spam from legitimate messages. By automating the detection of spam, this application contributes to reducing cybercrimes and simplifying the digital lives of users.

The project leverages Naive Bayes Classifiers, a family of algorithms based on Bayes’ Theorem, which presumes independence between predictive features. This theorem is crucial for calculating the likelihood of a message being spam based on various characteristics of the data.

Screenshots

Here are some screenshots of the application:

Home Page of the Spam Classifier

Displaying a not spam classification result

Displaying a spam classification result

Deployed Website

The classifier is accessible online at Spam Classifier Web App. Users can test the functionality by submitting text to be classified in real-time.

Features

• Real-Time Spam Detection: Quickly classify whether messages are spam.
• User-Friendly Interface: Easy-to-use interface built with Streamlit.
• High Accuracy: Employs Naive Bayes Classifier for high reliability.
• Data Visualization: Integrates graphical representations of data analytics.
• Multi-Format Support: Capable of analyzing both emails and SMS.

Technologies Used

Programming Language:

• Python

Environment:

• Jupyter Notebook
• Pycharm

Modules:

1. Streamlit - For the graphical user interface.
2. numpy, pandas - For data pre-processing.
3. matplotlib, seaborn, wordcloud - To represent data through graphs.
4. nltk - To work with human language data.
5. sklearn - To build the machine learning model.
6. pickle - To export the efficient machine learning model.
7. time - To delay function calls.

Design Software:

• Figma

Installation

Prerequisites

Before installing the application, ensure you have the following installed on your machine:

• Python (3.8 or newer)
• pip (Python package installer)

Steps

To get the Spam Classifier up and running locally on your machine, please follow the steps below:

1. Clone the repository:

   git clone https://github.com/yourusername/email-sms-spam-classifier.git
   cd email-sms-spam-classifier
   

2. Install required Python packages:

   pip install -r requirements.txt
   

3. Run the Streamlit application:

   streamlit run app.py
   

Contributions

Contributions are welcome! We value your input and appreciate your help in making the app even better.

License

This project is covered under the MIT License. See the LICENSE file for more information.