Spam Email Detection using CNN

📌 Overview

This project implements a Spam Email Detection system using a Convolutional Neural Network (CNN). The model classifies emails as either Spam or Ham (Not Spam) based on their content. Deep learning techniques are applied to enhance accuracy compared to traditional machine learning approaches.

📂 Dataset

The dataset used for training and evaluation consists of labeled emails categorized as spam or ham. It is preprocessed to extract text features and converted into numerical representations suitable for CNN processing.

Data Preprocessing

• Text Cleaning: Removal of special characters, stopwords, and extra spaces.
• Tokenization: Conversion of text into sequences.
• Padding: Ensuring uniform input size for CNN.
• Embedding Layer: Word embeddings for improved feature extraction.

🔧 Technologies Used

• Python
• TensorFlow / Keras
• Natural Language Processing (NLP)
• CNN for Text Classification
• Jupyter Notebook / Google Colab

🏗 Model Architecture

The CNN model is structured as follows:

1. Embedding Layer - Converts words into dense vectors.
2. Convolutional Layers - Extracts spatial features from text sequences.
3. Max-Pooling Layers - Reduces dimensionality and retains important information.
4. Fully Connected Layers - Final decision-making layers using dense layers.
5. Softmax Activation - Outputs probability for spam/ham classification.

🚀 Installation & Usage

1️⃣ Clone the Repository

git clone https://github.com/ayushk028/spam-email-detection-cnn.git
cd spam-email-detection-cnn

2️⃣ Install Dependencies

pip install -r requirements.txt

3️⃣ Train the Model

python train.py

4️⃣ Test the Model

python test.py

📈 Performance & Evaluation

• Accuracy: Achieved an accuracy of ~90% on test data.
• Loss Function: Categorical Cross-Entropy.
• Optimization Algorithm: Adam Optimizer.
• Evaluation Metrics: Precision, Recall, F1-Score.

📊 Results

The model effectively differentiates between spam and non-spam emails, outperforming traditional ML models such as Naive Bayes and SVM in terms of accuracy and generalization.

🔥 Future Improvements

• Implement Bidirectional LSTMs to improve sequential understanding.
• Use pre-trained word embeddings (e.g., Word2Vec, GloVe) for better text representation.
• Deploy the model as a web service or API.

🤝 Contributing

Feel free to fork this repository and contribute! Pull requests are welcome.

📜 License

This project is open-source and available under the MIT License.

📬 Contact

For queries, reach out at ayushk028.github.io.