Query-Based Online Data Pricing

In the rapidly developing information age,various industries generate a large amount of data,which contains a lot of valuable information.If it can be reasonably mined and analyzed,this information can bring great commercial value to the enterprise.Therefore,the importance of data has gradually gained attention and recognition.Since consumers have various needs for various data,more and more data can be sold and purchased online.In order to promote data transaction activities,a reasonable pricing mechanism is needed to allocate prices for the transaction data.At present,the data providers' self-pricing method is generally adopted.After browsing and analyzing,the consumers decide whether to purchase or how to purchase.Although the pricing method is simple,the transaction volume of both parties is large;and the consumers can get data only from one data provider at a time,which cannot meet their flexible and versatile needs;and there may also be arbitrage situations.Therefore,a more flexible pricing method is proposed: query-based data pricing,the data providers upload the databases with the defined price points in advance to the market service platform,and the consumers only need to submit queries to the platform,and the platform can generate an objective function and the corresponding constraints of an integer linear programming for each query according to databases,price points,and queries.The result of the solution is the price of a query.This pricing method satisfies arbitrage-free axiom and discount-free axiom,and the calculated price is the optimal price,which provides convenience for both parties.However,for each query,the system needs to separately generate the corresponding constraints and solve the calculation.When queries are related,some of the constraints are repeated.When calculating the price,this part of the constraints are repeatedly calculated,which affects the query efficiency;in order to avoid repeating the charging of the same views in the solution results,the query records are saved in the system,which will cause storage overhead;and when databases are updated,the data providers need to define the update price,which increases their workload.Based on query-based data pricing,this paper proposes an improved pricing method: querybased online data pricing.We abstract the data pricing problem into an online problem.According to the characteristics of the online algorithm,constraints of previous arrival must be satisfied,and value of a variable cannot decrease.Therefore,the first improvement in this paper is to save the variables whose value is 1 in the solution,determine whether each constrain in the process of generating constraints for the next query can be satisfied,then only add unrestricted constraints to the solution queue.After this improvement,the repeated calculation of the same constraints and the repeated charging of the same views are avoided at the same time,which can reduce the computational overhead and improve the query efficiency;Considering the problem of storage overhead and database update,the second improvement of this paper is to propose a strategy of setting the deadline.There are two situations: consumers who are not sensitive to data update can register an account and choose a time limit.During the validity period,the database status will remain unchanged and the system will save the query records.It is convenient for consumers to get results quickly.After the expiration date,the system will clear all the records under the account;consumers who are sensitive to data update can choose a deadline when submitting a query.During this period,the system will save the records of the query,and the calculation of each query is performed in the latest state of the databases,so consumers can get the latest data in time.The system will automatically delete the expired record.This strategy can effectively reduce the storage overhead and respond to the update of databases,and the system can automatically calculate the update price for the data providers,which makes the entire pricing process easier.