Research On Pricing Method Of Personal Big Data

With the advent of the era of artificial intelligence,data has become the most valu-able asset.As an important resource and asset,how to share and trade data,realize its circulation and discover its value is a hot topic in the industry.The data cost is difficult to estimate and the market value of the data is difficult to assess due to the diversity and dynamics of application scenarios.At the same time,the complex correlation between the data makes the data market arbitrage behavior.How to design a reasonable data pricing mechanism to make full use of big data resources and achieve mutual benefit and win-win situation is a key technical problem for the sustainable and healthy development of data trading.Personal big data,as an important big data resource,has been collected and analyzed by many data agents and sold for profit without the knowledge of individ-ual users.How to protect the rights and interests of individuals'data owners,enable individuals to actively participate in data transactions,contribute high-quality valuable data and obtain due benefits,and enable data buyers to obtain valuable data,so as to promote the healthy development of personal big data transaction market is the key issue to be solved urgently.Data pricing has become a hot research topic in the field of computer and economics.At present,most of the data pricing related work in the database field focuses on struc-tured and relational general data,but there is little research on pricing personal big data.The quality personal big data is the main factor affecting the value of personal big data,and pricing based on the quality of personal big data enables data buyers to obtain valu-able data.Entropy measures the amount of information contained in personal big data,and pricing based on information entropy can accurately reflect the value of personal data.Minimum provenance is used to measure the contribution to query results,and pricing based on provenance makes personal big data pricing fair.Differential privacy is an effective way to protect personal big data,which can be effectively utilized by making reasonable pricing and compensation according to the degree of privacy loss.The pricing mechanism of personal big data plays a very important role in the framework of personal big data trading market.This paper makes a systematic and in-depth study on the pricing method of personal big data from multiple perspectives.The main work and innovations are as follows:(1)We first study the information entropy pricing problem.Shannon's information entropy is used to measure the value contained in data tuples.Different data contain dif-ferent information entropy,different value,and different price.A new data pricing metric,data information entropy,is proposed to price personal big data from the perspective of information content,establish the functional relationship between information entropy and price,and realize the mapping of information entropy to price through the connec-tion function.The experimental results show that the pricing measurement method based on information entropy has a certain enlightening effect on the research of data pricing mechanism and will further promote the development of personal data trading market.(2)We then study minimum provenance pricing and conducts the pricing analysis based on minimum provenance query according to the price setting function of source tuple and value weight.The importance of the data itself(intrinsic quality)and the cor-relation between the data are taken into account.Firstly,a provenance pricing method for personal big data is proposed,which is based on the minimum provenance that con-tributes to the query results.Then the exact algorithm and approximate algorithm are designed to calculate the exact price and approximate price of the query respectively.In the end,the selection-join query and complex query instantiation pricing are used to validate the two real datasets,and their performance is evaluated extensively.The experimental results show that the pricing method is effective and efficient.(3)We finally study differential privacy pricing and analyzes how to fully exploit the value of personal big data while protecting personal privacy.We propose a privacy pricing method for personal big data,design two different mechanisms of positive pricing and reverse pricing to reasonably price private data,and compensates appropriately according to the degree of privacy loss of individual users,so as to realize the privacy protection and reasonable utilization of personal big data.By adding noise to personal big data and desensitizing data,different degrees of privacy protection can be achieved to obtain data with different accuracy.According to different privacy protection needs of individuals,different noises are added to obtain data with different accuracy,and differential pricing is realized.The experimental results show that this pricing method can reasonably price the personal data,fairly compensate the loss of personal privacy,ensure the practicability of data,and achieve a balance between privacy protection and data utility.At present,the research on the pricing method of personal big data by domestic and foreign research institutions is still in the exploratory stage,and there are many problems such as fairness,rationality,practicability and arbitrage-free.In this paper,the personal big data pricing method is studied and explored from the aspects of information entropy,data traceability and differential privacy,which makes a beneficial contribution to the research of personal big data trading and pricing method.