Model Card for Infinitode/SMCM-OPEN-ARC

Repository: https://github.com/Infinitode/OPEN-ARC/

Model Description

OPEN-ARC-SMC is a MultinomialNB model developed as part of Infinitode's OPEN-ARC initiative. It was created to categorize text, particularly emails, as either spam or legitimate (ham).

Architecture:

• MultinomialNB: Used default parameters.
• Framework: SKLearn.
• Training Setup: Trained using default params.

Uses

• Determining whether emails or SMS are spam or legitimate.
• Enhancing research and developing defensive measures against spammers.

Limitations

Emails or SMS may be classified as false positives or false negatives because of the nature of the data and its inherent limitations.

Training Data

• Dataset: Spam Mail Classifier Dataset dataset from Kaggle.
• Source URL: https://www.kaggle.com/datasets/mosapabdelghany/spam-mail-classifier/
• Content: Messages categorized as either spam or ham (legitimate emails or SMS).
• Size: 1000 email/SMS messages labeled as spam or ham.
• Preprocessing: The preprocessing steps included removing missing values and converting text into vectors.

Training Procedure

• Metrics: accuracy, precision, recall, F1
• Train/Testing Split: 80% train, 20% testing.

Evaluation Results

Metric	Value
Testing Accuracy	98.48%
Testing Precision (spam)	96.15%
Testing Recall (spam)	93.17%
Testing F1 (spam)	94.64%

How to Use

new_emails = [
    "Congratulations! You've won a free prize. Click the link to claim.", # Likely spam
    "Hi, just confirming our meeting for tomorrow at 10 AM. Thanks." # Likely not spam
]

# Vectorize the new emails using the fitted vectorizer
new_emails_vectorized = vectorizer.transform(new_emails)

# Make predictions
predictions = model.predict(new_emails_vectorized)

for i, email in enumerate(new_emails):
    print(f"\nEmail: '{email}'")
    print(f"Prediction: {predictions[i]}")

Contact

For questions or issues, open a GitHub issue or reach out at https://infinitode.netlify.app/forms/contact.