Research On Pricing Mechanism In Data Transcation

In this paper,based on the existing data pricing algorithm does not consider the impact of privacy,there is no privacy measurement method based on transaction data set,and the existing data pricing method can not separately price the tuple,two privacy metrics for transaction data sets and a method based on tuple-based data pricing are proposed.Finally,the rationality of the theory is proved by experiments.The specific work of this paper and the results obtained are as follows:(1)"privacy level","data set description object"and a privacy measurement method based on information entropy are proposed.According to the description content of the transaction data set and Shannon information theory,the attributes of the transaction data set can be classified,and the privacy content of the transaction data set and each tuple can be calculated.(2)A privacy metric method based on set pair analysis is proposed.The information entropy-based privacy metric proposed in the previous article can accurately measure the privacy content of the transaction data set.However,data trading platforms tend to desensitize data during the process of data purchases,the set pair analysis theory is used to remove some sensitive information in the transaction data set.With combining the attribute information,calculating meta-group relative information leakage set entropy and the relative information leakage set scoring function entropy,the current privacy content of the tuple can be obtained..(3)A tuple pricing method with privacy metric is proposed.Based on the privacy metric calculated by the privacy metric method proposed in this paper,combined with the data reference index,a dual group is constructed as a factor affecting the tuple price,and each factor is assigned a weight.With calculating the ratio of the factors in each tuple to the entire transaction data set factor and combining it with the weight,the price of each tuple in the data set can be calculated based on the price of the entire data set.(4)A data pricing simulation system is coded to verify the pricing mechanism based on experiments.the transaction data set is obtained by the web crawler.The experimental results show that the privacy metric proposed in this paper can better calculate the privacy content of each tuple in the dataset,and the tuple pricing method proposed in this paper has higher pricing accuracy than the traditional tuple pricing method and the lower time complexity.