Research On Data Pricing Methods Based On VCG And Rubinstein Models

With the rapid development of networks and information technology,technologies and applications such as social networks and the Internet of Things are generating massive amounts of data every day.Faced with such a large amount of data,data storage and calculation are no longer the only purpose.How to effectively dig out huge commercial value from these massive data has become the focus of all parties' attention.The current research results have found that although the scale of data is increasing,there are "data islands",and the existence of data islands seriously hinders the mining of data value.Therefore,sufficient data circulation and sharing has become the key to solving the problem.Data transaction as an incentive mechanism for data sharing has been extensively researched and developed in recent years,but there are still some problems in promoting data sharing.The most important one is that effective data pricing has not been resolved,which leads to the current data The trading market is still in its preliminary form.In order to solve the problem of data pricing,this paper proposes two pricing mechanisms: auction-based data transaction pricing mechanism and data transaction pricing mechanism based on Rubinstein model.In terms of the auction-based data pricing mechanism,this article considers the situation of a single data owner and multiple data consumers,and takes into account the differences in data consumers' demand for similar data,and designs a data scale based on Data division method of layer auction.This data division method also solves the problem that the data owner cannot determine the amount of auction data copied,that is,the data owner cannot know how many copies of the same data should be copied for sale in order to maximize their own interests.Since the auction proposed in this article takes place in the Internet scenario,compared with traditional live auctions,this auction method has the problem of false identity bidding attacks,that is,data consumers can participate in the auction process by registering multiple accounts.The collusion between them can influence the auction results,thereby increasing their own profits.This paper theoretically proves that the pricing mechanism has the characteristics of preventing false identity bidding attacks,that is,under the action of this mechanism,even if data consumers conduct false identity bidding attacks,they cannot get higher profits,so they have no behavioral motivation.In terms of data transaction pricing mechanism based on Rubinstein's model,this article considers the situation of a single data consumer and multiple data owners.Data consumers can subscribe to data according to their own needs,and then form a Rubinstein bargaining game with multiple data owners.This paper calculates the best bidding strategy for the game participants in the data transaction scenario.