Efficient and Privacy-Preserving Binary Dot Product via Multi-Party Computation

2025-10-17

AI Summary

Binary vector dot-product computation in tree-structured vertical federated learning faces a fundamental privacy-efficiency trade-off. Method: This paper proposes BiMPC—a secure multi-party computation framework comprising (i) DoMA, a modular-addition-based binary dot-product protocol enabling bit-level fine-grained privacy; (ii) integration of high-domain random linear masking with three-party oblivious transfer (3-OT) to circumvent the inefficiency of Shamir secret sharing for bitwise operations; and (iii) a multi-party collaborative computing architecture optimizing communication and computational overhead. Contribution/Results: Evaluated experimentally, BiMPC achieves up to 3.2× higher dot-product throughput than state-of-the-art approaches while maintaining rigorous security guarantees under standard cryptographic assumptions. It scales to thousands of binary features and hundreds of participating nodes, significantly advancing the practicality of binary-data vertical federated learning.

Technology Category

• Machine Learning: Privacy
• Computer Vision: Bias, Fairness & Privacy
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Security and Privacy: Privacy-enhancing technologies
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Striking a balance between protecting data privacy and enabling collaborative computation is a critical challenge for distributed machine learning. While privacy-preserving techniques for federated learning have been extensively developed, methods for scenarios involving bitwise operations, such as tree-based vertical federated learning (VFL), are still underexplored. Traditional mechanisms, including Shamir's secret sharing and multi-party computation (MPC), are not optimized for bitwise operations over binary data, particularly in settings where each participant holds a different part of the binary vector. This paper addresses the limitations of existing methods by proposing a novel binary multi-party computation (BiMPC) framework. The BiMPC mechanism facilitates privacy-preserving bitwise operations, with a particular focus on dot product computations of binary vectors, ensuring the privacy of each individual bit. The core of BiMPC is a novel approach called Dot Product via Modular Addition (DoMA), which uses regular and modular additions for efficient binary dot product calculation. To ensure privacy, BiMPC uses random masking in a higher field for linear computations and a three-party oblivious transfer (triot) protocol for non-linear binary operations. The privacy guarantees of the BiMPC framework are rigorously analyzed, demonstrating its efficiency and scalability in distributed settings.

Problem

• Optimizing multi-party computation for binary vector dot products
• Enhancing privacy in bitwise operations for vertical federated learning
• Addressing inefficiency in existing methods for binary data computation

Innovation

• BiMPC framework for binary multi-party computation
• DoMA method using modular addition for dot product
• Random masking and triot protocol for privacy