Self-Reinforcing Controllable Synthesis of Rare Relational Data via Bayesian Calibration

Imbalanced data is commonly present in real-world applications. While data synthesis can effectively mitigate the data scarcity problem of rare-classes, and LLMs have revolutionized text generation, the application of LLMs to relational/structured tabular data synthesis remains underexplored. Moreover, existing approaches lack an effective feedback mechanism that can guide LLMs towards continuously optimizing the quality of the generated data throughout the synthesis process. In this work, we propose RDDG, Relational Data generator with Dynamic Guidance, which is a unified in-context learning framework that employs progressive chain-of-thought (CoT) steps to generate tabular data for enhancing downstream imbalanced classification performance. RDDG first uses core set selection to identify representative samples from the original data, then utilizes in-context learning to discover the inherent patterns and correlations among attributes within the core set, and subsequently generates tabular data while preserving the aforementioned constraints. More importantly, it incorporates a self-reinforcing feedback mechanism that provides automatic assessments on the quality of the generated data, enabling continuous quality optimization throughout the generation process. Experimental results on multiple real and synthetic datasets demonstrate that RDDG outperforms existing approaches in both data fidelity and downstream imbalanced classification performance. We make our code available at https://github.com/cszhangLMU/RDDG.