电子学报2025，Vol.53Issue(9)：3089-3102,14.DOI:10.12263/DZXB.20250451
面向LLM开放域问答中多方私有表格筛选:一种MPC可公开聚合审计与动态信誉的增强方法
Multi-Party Private Table Screening for LLM-Driven ODQA:An Enhanced Method with MPC,Publicly Aggregable Audit,and Dynamic Reputation
摘要
Abstract
Large language model(LLM)driven open-domain question answering(ODQA)systems,exemplified by frameworks like GIST(Generating Identifiers and Selecting chunks for Tables),have garnered considerable research atten-tion due to their significant potential in processing extensive tabular data.However,when such ODQA systems integrate da-ta from multiple providers for Top-K candidate screening,traditional methods requiring access to raw data encounter sub-stantial challenges concerning data privacy,computational transparency,and participant trustworthiness.While existing re-search employs zero-knowledge proofs and stake-based mechanisms to achieve public verifiability,the overhead of generat-ing and verifying individual proofs in large-scale scenarios is often prohibitive.Moreover,conventional stake-based mecha-nisms exhibit limitations in fairness and adaptability within dynamic environments.This paper proposes an enhanced meth-od for multi-party private table screening in LLM-driven ODQA,which integrates multi-party computation(MPC),a public-ly aggregable audit mechanism,and a dynamic reputation system.This study adapt the Top-K multi-party private table screening process using MPC to ensure data privacy.Concurrently,an efficient aggregable audit mechanism is introduced;this mechanism combines zero-knowledge proof techniques with random sampling,aggregate proof construction,time-win-dow-based batching,and error localization,thereby enabling the public and batch-verified correctness of the scoring and ranking process.The integration of a blockchain-based dynamic reputation feedback mechanism further enhances system fairness and constrains malicious behavior.Experimental evaluations demonstrate that our Top-K candidate screening meth-od,while preserving privacy,achieves high consistency with the original GIST screening approach,attaining a Top-50 aver-age recall of 0.91 and an average Jaccard index of 0.83,thus indicating minimal impact on end-to-end ODQA task perfor-mance.Furthermore,the efficiency of publicly auditable proof generation and verification for large-scale tasks is significant-ly improved,saving approximately 87%of proof time compared to individual proofs.The adaptability and fairness of the feedback mechanism are also demonstrably enhanced.关键词
开放域问答/大语言模型/多方安全计算/可公开审计/零知识证明/区块链Key words
open-domain question answering/large language models/secure multi-party computation/publicly au-ditable/zero-knowledge proofs/blockchain分类
信息技术与安全科学引用本文复制引用
胡睿,吴昊,潘宇轩,张琳,刘雨,朱孔林..面向LLM开放域问答中多方私有表格筛选:一种MPC可公开聚合审计与动态信誉的增强方法[J].电子学报,2025,53(9):3089-3102,14.基金项目
国家重点研发计划(No.2023YFB2704500) National Key Research and Development Program of China(No.2023YFB2704500) （No.2023YFB2704500）
