Privacy-Preserving Synthetic Review Generation with Diverse Writing Styles Using LLMs

TL;DR

This paper evaluates the diversity and privacy of synthetic reviews generated by LLMs, identifies their limitations, and proposes a prompt-based method to improve diversity while maintaining privacy.

Contribution

It introduces comprehensive metrics for assessing synthetic review diversity and privacy, and proposes a prompt-based approach to enhance diversity without compromising privacy.

Findings

LLMs have significant limitations in generating diverse synthetic reviews.

Current synthetic data poses privacy risks such as re-identification.

Prompt-based techniques can improve diversity while preserving privacy.

Abstract

The increasing use of synthetic data generated by Large Language Models (LLMs) presents both opportunities and challenges in data-driven applications. While synthetic data provides a cost-effective, scalable alternative to real-world data to facilitate model training, its diversity and privacy risks remain underexplored. Focusing on text-based synthetic data, we propose a comprehensive set of metrics to quantitatively assess the diversity (i.e., linguistic expression, sentiment, and user perspective), and privacy (i.e., re-identification risk and stylistic outliers) of synthetic datasets generated by several state-of-the-art LLMs. Experiment results reveal significant limitations in LLMs' capabilities in generating diverse and privacy-preserving synthetic data. Guided by the evaluation results, a prompt-based approach is proposed to enhance the diversity of synthetic reviews while preserving reviewer privacy.