This research tackles the critical challenge of fake review detection across multiple languages and domains using data augmentation techniques with large language models.

• Addresses the lack of training data for low-resource languages and domains
• Utilizes LLMs to generate synthetic training examples
• Enhances detection capabilities across linguistic and commercial boundaries
• Improves security for e-commerce platforms through better fraud detection

For security professionals, this work provides valuable techniques to protect platform integrity and consumer trust by identifying deceptive content at scale, particularly in markets where training data is scarce.

Data Augmentation for Fake Reviews Detection in Multiple Languages and Multiple Domains