Privacy-Preserving Data Deduplication for Enhancing Federated Learning of Language Models (Extended Version)

TL;DR

This paper introduces EP-MPD, a privacy-preserving deduplication protocol for federated learning of language models, significantly improving efficiency and model performance while maintaining data privacy.

Contribution

The paper presents a novel, modular protocol for privacy-preserving deduplication in federated learning using two new Private Set Intersection variants.

Findings

Up to 19.62% improvement in perplexity

Up to 27.95% reduction in running time

Effective privacy-performance balance in large-scale federated learning

Abstract

Deduplication is a vital preprocessing step that enhances machine learning model performance and saves training time and energy. However, enhancing federated learning through deduplication poses challenges, especially regarding scalability and potential privacy violations if deduplication involves sharing all clients' data. In this paper, we address the problem of deduplication in a federated setup by introducing a pioneering protocol, Efficient Privacy-Preserving Multi-Party Deduplication (EP-MPD). It efficiently removes duplicates from multiple clients' datasets without compromising data privacy. EP-MPD is constructed in a modular fashion, utilizing two novel variants of the Private Set Intersection protocol. Our extensive experiments demonstrate the significant benefits of deduplication in federated learning of large language models. For instance, we observe up to 19.62\% improvement in perplexity and up to 27.95\% reduction in running time while varying the duplication level between 10\% and 30\%. EP-MPD effectively balances privacy and performance in federated learning, making it a valuable solution for large-scale applications.