Title :
Information-Theoretically Secure Number-Product Protocol
Author :
Shen, Chih-Hao ; Zhan, Justin ; Wang, Da-Wei ; Hsu, Tsan-Sheng ; Liau, Churn-Jung
Author_Institution :
Acad. Sinica, Taipei
Abstract :
A secure multiparty number-product protocol is an important building block in the area of secure multiparty computation. With proper composition of the building block, most applications, such as circuit evaluation, data mining, and private information retrieval, can be executed securely and collaboratively by potentially dishonest parties. In this work, we propose a commodity-based secure number-product protocol, the security of which has been validated with the metrics proposed by Chiang et al. based on information theory. We prove that the protocol can be executed securely under the assumption of semi-honest behavior. Furthermore, if an extra semi-trusted party exists, the proposed protocol can resist any number of corrupt parties. In addition to the security issue, we compare the protocol´s communication and computation costs with those of the theoretically secure solution proposed by Ben-Or et al.
Keywords :
information theory; protocols; security of data; commodity; information theory; multiparty number-product protocol security; Circuits; Collaborative work; Computational efficiency; Data mining; Data security; Information retrieval; Information security; Information theory; Protocols; Resists; Commodity-based Paradigm; Number-product Protocol; Secure Multiparty Computation;
Conference_Titel :
Machine Learning and Cybernetics, 2007 International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-0973-0
Electronic_ISBN :
978-1-4244-0973-0
DOI :
10.1109/ICMLC.2007.4370663