imabedalghafer/arabic-ner-masking

This model is a fine-tuned version of a BERT model for Named Entity Recognition (NER) specifically designed to identify and mask sensitive information in Arabic text.

Model Description

This model can identify entities that should be masked for privacy protection, such as:

• Personal names
• Email addresses
• Company names
• Other sensitive information

Usage

from transformers import pipeline

# Load the model
ner_pipeline = pipeline("ner", model="imabedalghafer/arabic-ner-masking")

# Example usage
text = "مرحبا يا أحمد، يمكنك إرسال البريد إلى [email protected]"
results = ner_pipeline(text)
print(results)

Training Data

The model was fine-tuned on Arabic text data with manually annotated entities for masking.

Performance

The model achieves good performance on entity detection and masking tasks. See the evaluation metrics in the training logs.

Intended Use

This model is intended for:

• Privacy protection in Arabic text
• Automated data anonymization
• Content filtering and moderation

Limitations

• Performance may vary on out-of-domain text
• May not capture all types of sensitive information
• Requires post-processing for complete anonymization