Augmenting Anonymized Data with AI: Exploring the Feasibility and Limitations of Large Language Models in Data Enrichment

📅 2025-04-03
📈 Citations: 1
Influential: 0
📄 PDF
🤖 AI Summary
Privacy-preserving techniques such as k-anonymity often induce data sparsity, information loss, and degraded downstream modeling performance. Method: This paper proposes a large language model (LLM)-based approach for anonymized data augmentation, centered on a privacy-constraint-aware, customized prompting strategy that rigorously enforces k-anonymity, l-diversity, and t-closeness requirements. The method integrates structured prompt templates, the pyCanon privacy verification framework, and real-world datasets to enable verifiable, privacy-compliant data expansion. Contribution/Results: We present the first systematic evaluation of LLMs’ capability to generate privacy-compliant synthetic data. Experimental results demonstrate that the augmented data significantly enhances feature richness and improves downstream prediction accuracy by an average of +8.2%, without violating anonymity constraints—establishing a novel paradigm for jointly optimizing privacy protection and model utility.

Technology Category

Application Category

📝 Abstract
Large Language Models (LLMs) have demonstrated advanced capabilities in both text generation and comprehension, and their application to data archives might facilitate the privatization of sensitive information about the data subjects. In fact, the information contained in data often includes sensitive and personally identifiable details. This data, if not safeguarded, may bring privacy risks in terms of both disclosure and identification. Furthermore, the application of anonymisation techniques, such as k-anonymity, can lead to a significant reduction in the amount of data within data sources, which may reduce the efficacy of predictive processes. In our study, we investigate the capabilities offered by LLMs to enrich anonymized data sources without affecting their anonymity. To this end, we designed new ad-hoc prompt template engineering strategies to perform anonymized Data Augmentation and assess the effectiveness of LLM-based approaches in providing anonymized data. To validate the anonymization guarantees provided by LLMs, we exploited the pyCanon library, designed to assess the values of the parameters associated with the most common privacy-preserving techniques via anonymization. Our experiments conducted on real-world datasets demonstrate that LLMs yield promising results for this goal.
Problem

Research questions and friction points this paper is trying to address.

Exploring LLMs for enriching anonymized data without compromising privacy
Assessing effectiveness of LLM-based anonymized data augmentation techniques
Validating anonymization guarantees using pyCanon library on real datasets
Innovation

Methods, ideas, or system contributions that make the work stand out.

Using LLMs for anonymized data enrichment
Ad-hoc prompt templates for data augmentation
Validating anonymity with pyCanon library
S
Stefano Cirillo
Department of Computer Science, University of Salerno
D
Domenico Desiato
Department of Computer Science, University of Bari Aldo Moro
G
Giuseppe Polese
Department of Computer Science, University of Salerno
M
Monica Maria Lucia Sebillo
Department of Computer Science, University of Salerno
G
Giandomenico Solimando
Department of Computer Science, University of Salerno