Abstract: The cross-organizational data has the typical characteristics of big data in collaborative optimization decision-making, such as distributedness, heterogeneity, and privacy etc. Secure multiparty computation(SMC) is a privacy-preserving algorithm, based on collaborative mechanisms or protocols. However, the methods typically used, such as monotone span program, cannot get rid of the computational complexity. It discussed two kinds of   problems in collaborative optimization decision, and proposed secure multi-party computation protocols for the privacy-preserving of constraint parameters and decision variables. Then, it gave security proof. The results show that the SMC protocols can reduce the computational complexity of collaborative optimization decisions, and computation of some private information can be completed without transferred processing.

Key words: big data, collaborative optimization, privacy-preserving, secure multi-party computation