Research On Pricing Of Big Data Information Service From The Perspective Of Uncertainty Analysis

As increasing application of big data,more and more companies are investing in the field of big data analytics and applications,providing customers with a variety of big data information services.For example,the in-vehicle information system can record the dynamic information of the car,such as the speed of the car,the mileage,and the presence or absence of an emergency braking system.The insurance company evaluates the risk level of the customer based on the information and designs a personalized premium plan.When big data information services are provided to the outside world,how to price big data information services has become a focus issue for both the supplier and demander.In view of this,this paper takes the pricing problem of big data information service as object of the research.The main contents are:(1)Point out the problems existing in the pricing of existing big data information services,and summarize the value transmission mechanism of big data information services.The comparison of commodity types in the merchandise trichotomy shows that big data information services are exploratory commodities.On the basis of analyzing the hierarchical structure of big data information service and the influencing factors of pricing,the value chain model of big data information service is summarized.It is pointed out that the quality effect of big data information service requires multi-level delivery confirmation of downstream customers,which has a high degree.Uncertainty,traditional pricing methods are difficult to use,and result-oriented pricing—the royalty-based pricing method is more reasonable.In the upstream of the big data information service value chain,big data information service providers need to use multi-channel and multi?level data,and the provider needs to compensate the original data source to protect the legitimate rights and interests of the original data source.Infringement.(2)Propose a pricing model for data-type(data sets)big data information services.Based on the analysis of the uncertainty of the data set,a method to deal with the uncertainty of the data set is proposed.The data set pricing model is constructed.Firstly,the ranking of the five attributes of the data set is determined by the Delphi method.Then the analytic hierarchy process is used to determine the attribute weight and construct the data intrinsic value model.The value of the "maintenance data set,of J Auto Insurance Service Company was measured,and the calculation results showed the validity of the pricing model.This pricing method uses the experience of experts to circumvent the arbitrariness of attribute weight setting,and uses Monte Carlo simulation to avoid the uncertainty caused by the small number of experts.(3)Research on the pricing methods of different stages of solution-based big data information services.Based on the analysis of the uncertainty of different life cycle stages,the pricing system of solution-type big data information service is proposed.The focus is on the result-oriented pricing method—the combination of the royalty-like method and the guarantee mechanism based on smart contracts.The use of the smart contract guarantee mechanism ensures that the revenue of the big data information service demander is traceable and non-repudiation,and guarantees the rights of the big data information service provider.Eliminate the uncertainty of the quality effect between the supply and demand sides.(4)Define the ownership of big data information service,study the value of social network users and data value,propose a pricing model that considers the compensation of original data source,and design the structure and data of revenue compensation platform based on blockchain.The transaction process and technically traceable data.(5)Based on the empirical practice of big data information service pricing of J Auto Insurance Service Company,the effectiveness of the solution-based big data information service pricing system was preliminarily verified.Introducing a result-oriented pricing method in the growth period,developed a "blockchain-based trading system",and established an "observation account" for big data information service providers to observe all transaction information on the nodes.It can guarantee the traceability and non-repudiation of buyer transactions.According to the transaction data actually generated by the company,the various attribute values(ie,price influencing factors)of the customer(auto insurance company)are described as the input neurons of the BP neural network,and the transaction price is used as the output neuron.After training and testing,the test value is highly compatible with the real value and is an effective pricing method.On the basis of summarizing the research results,the paper finally looks forward to the future research direction.The main innovations of this paper are:(1)To deal with the uncertainty factors of data-type big data information service itself,the analytic hierarchy process and Monte Carlo simulation are combined to propose a data-based big data information service pricing method.The weight of each attribute is determined by AHP.Monte Carlo simulation is used to give the attribute value of the data set in the instance,and the value of the instance data set is measured.The model avoids the arbitrariness of weight setting and the uncertainty caused by too little data.(2)To deal with the uncertainty of the quality effect of solution-type big data information service,this paper proposes a result-oriented type-based pricing method and introduces smart contract as its guarantee mechanism,which simulates the traceability and non-repudiation of transactions.(3)In response to the frequent occurrence of data leakage incidents,the author proposes a pricing model that considers the user(original data source)compensation,and uses the blockchain technology to achieve data traceability,ensuring that each data user conducts the original data source.Compensation avoids the risk of data users infringing privacy legally and compensates the legitimate rights and interests of the original data source.(4)To deal with uncertainty of the relationship between transaction price and influencing factors,this paper uses the actual data of J auto insurance service company big data information service transaction,and uses BP neural network to train the relationship between historical transaction price and influencing factors.The results show the effectiveness of BP neural network for the pricing of big data information services.The contribution of the BP neural network pricing method is that,in the mature period of big data information service,the mapping relationship between the price and the influencing factors of big data information service is trained according to the historical transaction record.The later pricing can be directly given by the attribute value of the customer.In theory,the "one-on-one" bargaining method in reality can be ended.Improve transaction efficiency and provide new ideas for pricing big data information services.