创造始于理解：将大型语言模型（LLM）用作隐私保护型表格数据合成的策略设计者

中文摘要 AI 辅助

高风险领域的表格数据共享受隐私法规约束，合成数据是颇具前景的替代方案，但深度生成模型训练成本高且难以审计，而基于大型语言模型（LLM）的方法常将记录序列化为文本，模糊表格结构并暴露敏感数据。我们提出表格合成策略设计者（Tabular Synthesis Strategy Designer，TabSSD），该方法利用LLM设计合成流程而非直接生成记录。TabSSD向LLM提供变量依赖的树衍生摘要而非原始记录，从而生成用于本地执行与评估的Python程序。在12个数据集上，TabSSD在统计保真度、预测效用与经验隐私风险间取得良好平衡，在6项指标上于10种方法中获得最优平均排名；且相较对比方法，其大幅降低了本地计算量与令牌消耗。通过支持人工引导优化并消除用户侧模型调优，TabSSD降低了透明表格数据合成所需的专业知识与基础设施门槛。

英文摘要

Sharing tabular data in high-stakes domains is constrained by privacy regulations. Synthetic data offer a promising alternative, but deep generative models are costly to train and difficult to audit, while LLM-based methods often serialize records as text, obscuring tabular structure and exposing sensitive data. We introduce Tabular Synthesis Strategy Designer (TabSSD), which uses an LLM to design synthesis procedures rather than directly generate records. TabSSD provides the LLM with tree-derived summaries of variable dependence rather than raw records, which produces Python programs for local execution and evaluation. Across twelve datasets, TabSSD strikes a favourable balance among statistical fidelity, predictive utility, and empirical privacy risk, achieving the best average rank across six metrics among ten methods. Moreover, it substantially reduces local computation and token consumption relative to the compared methods. By enabling human-guided refinement and eliminating user-side model tuning, TabSSD lowers the expertise and infrastructure barriers to transparent tabular data synthesis.

发表机构

• School of Artificial Intelligence, Jilin University（吉林大学人工智能学院）
• Broad College of Business, Michigan State University（密歇根州立大学broad商学院）
• Commonwealth Scientific and Industrial Research Organisation (CSIRO)（英联邦科学与工业研究组织）
• Center of Excellence for Generative AI, King Abdullah University of Science and Technology (KAUST)（阿卜杜拉国王科技大学生成式人工智能卓越中心）
• The Swiss AI Lab, IDSIA-USI/SUPSI（瑞士人工智能实验室IDSIA-USI/SUPSI）

机构由 AI 辅助整理，请以论文原文为准。