Group-wise Prompting for Synthetic Tabular Data Generation using Large Language Models

Generating realistic synthetic tabular data presents a critical challenge in machine learning. This study introduces a simple yet effective method employing Large Language Models (LLMs) tailored to generate synthetic data, specifically addressing data imbalance problems. We propose a novel group-wise prompting method in CSV-style formatting that leverages the in-context learning capabilities of LLMs to produce data that closely adheres to the specified requirements and characteristics of the target dataset. Moreover, our proposed random word replacement strategy significantly improves the handling of monotonous categorical values, enhancing the accuracy and representativeness of the synthetic data. The effectiveness of our method is extensively validated across eight real-world public datasets, achieving state-of-the-art performance in downstream classification and regression tasks while maintaining inter-feature correlations and improving token efficiency over existing approaches. This advancement significantly contributes to addressing the key challenges of machine learning applications, particularly in the context of tabular data generation and handling class imbalance. The source code for our work is available at: https://github.com/seharanul17/synthetic-tabular-LLM