TWI664598B - Financial service method with point of sale terminal - Google Patents

Abstract

A financial service method includes the following steps: (A) when receiving product parameter data related to a plurality of products related to a point-of-sale terminal, a financial servo unit stores the product parameter data into product management data; and (B) when receiving at least one financial service request from the point-of-sale terminal, the financial server unit uses a type of neural network algorithm based on the product parameter data of a current time interval and the hidden state value calculated in the previous time interval, Predict the hidden status value of the current time interval, and calculate a credit rating related to the point of sale terminal based on the hidden status value of the current time interval; and (C) the financial server unit generates and transmits the credit rating based on the credit rating At least one pair of financial service conditions that should correspond to at least one financial service request to the point of sale terminal.

Description

Financial service method with point of sale terminal

The invention relates to a financial service method, in particular to a financial service method for providing financial services by using a point-of-sale terminal vehicle.

Small and small profit enterprises usually refer to self-employed (including unpaid family employees) and self-employed small businesses. Although small and low-profit enterprises usually do not need too high start-up capital, the vast majority of small and small-profit enterprises are still vulnerable to shortage of funds. Compared with small and medium-sized enterprises, the risk cost and credit cost of small and meager corporate loans are higher, so general commercial banks are reluctant to get involved in such businesses. In addition, micro-profit companies usually do not have complete financial and financial statement records, and it is difficult for banks to verify their operating conditions. Considering risk control, it is often difficult for micro-profit companies to make loans.

However, the establishment and development of small and meager enterprises have a positive effect on creating a large number of self-employment opportunities, assisting vulnerable groups, promoting economic development and maintaining social stability. Therefore, for the bankers, how to support the establishment and development of small and meager enterprises and reduce the bank's lending risk is an urgent problem to be solved.

Therefore, an object of the present invention is to provide a financial service method for providing customized financial services and reducing bank risks.

Therefore, the financial service method of the present invention is implemented by a point-of-sale terminal. The point-of-sale terminal can provide an enterprise resource planning function and is connected to a management terminal and a financial server unit through a communication network. The communication network is connected to the financial server unit. The financial server unit stores user data related to the point of sale terminal. The user data includes product management data. The financial service method includes the following steps:

(A) the point-of-sale terminal generates a product management record including a product purchase record, a product sales record, and a product inventory record related to a plurality of products through a user's input operation within a time interval;

(B) the point-of-sale terminal obtains the corresponding time interval based on the product management record and includes at least one purchase parameter related to the purchase of the products, at least one sales parameter related to the sale of the products, and At least one product parameter data of inventory parameters related to the inventory of the products, and transmitting the product parameter data to the management terminal, so that the management terminal transmits the product parameter data to the financial server unit;

(C) The point-of-sale terminal generates and transmits at least one financial service request to the financial server unit through an input operation of the user, so that the financial server unit may compare one of the stored product management data with the current time point The product parameter data of the shortest time interval is used as the product parameter data of the current time interval, and a type of neural network algorithm is used to calculate at least based on the product parameter data of the current time interval and the previous time interval and is related to a type of neural network. The hidden state value of the hidden layer of the road model, predicts a hidden state value of the current time interval that is related to the hidden layer of the neural network model, and calculates a related to the sales based on the hidden state value of the current time interval. A credit rating of a point terminal, and generating at least one pair of financial service conditions that should correspond to at least one financial service request according to the credit rating; and

(D) When the point of sale terminal receives the at least one financial service condition generated by the financial servo unit, the point of sale terminal presents the at least one financial service condition.

Another object of the present invention is to provide a financial service method for providing customized financial services and reducing bank risks.

Therefore, the financial service method of the present invention is implemented by a financial server unit, which is connected to a management terminal and a point-of-sale terminal via a communication network. The point-of-sale terminal can provide an enterprise resource planning function. The server unit stores user data related to the point of sale terminal, the user data includes product management data, and the financial service method includes the following steps:

(A) When the financial servo unit receives a purchase parameter generated by the point-of-sale terminal that corresponds to a time interval and contains at least one related to the purchase of multiple products, at least one related to the sales of those products When the sales parameter of the product and the product parameter data of at least one inventory parameter related to the inventory of the products, the financial servo unit stores the product parameter data in the product management data;

(B) When the financial server unit receives at least one financial service request, the financial server unit uses the product parameter data of the time interval with the shortest interval between one of the stored product management data and the current time point as a current time Product parameter data in the interval, and use a type of neural network algorithm to predict the current state based on the product parameter data in the current time interval and the hidden state value of the hidden layer of a type of neural network model calculated in the previous time interval. A time interval related to the hidden state value of the hidden layer of the neural network model, and a credit rating related to the point of sale terminal is calculated based on the hidden state value of the current time interval; and

(C) The financial servo unit generates at least one pair of financial service conditions corresponding to at least one financial service request according to the credit rating, and transmits the at least one financial service condition to the point of sale terminal.

The effect of the present invention is that the product parameter data is transmitted to the management terminal through the point of sale terminal, so that the management terminal transmits the product parameter data to the financial server unit, so that the financial server unit makes the product parameter data Stored in the product management data, the credit generated by the financial servo unit based on the product parameter data of the current time interval and the hidden state value of the hidden layer of the neural network model, calculated from the previous time interval Ratings are generated in response to the operating conditions of customers, thereby providing customers with customized financial services and reducing the risk of banks from providing financial services such as lending.

Referring to FIG. 1, a financial service system includes an embodiment of a point-of-sale terminal 1 of the present invention, an embodiment of a financial server unit 2 of the present invention connected to the point-of-sale terminal 1 via a communication network 100, and a communication server 100 A management terminal 3 connected to the point of sale terminal 1 and the financial server unit 2.

The point of sale terminal 1 can provide a sales checkout function and an enterprise resource planning function, and includes a terminal input unit 11, a terminal display unit 12, a terminal communication unit 13 connected to the communication network 100, and an electrical connection The terminal input unit 11, the terminal display unit 12 and the terminal processing unit 14 of the terminal communication unit 13. The terminal processing unit 14 includes a sales management module 141 for providing the sales checkout function, an enterprise resource planning module 142 for providing the enterprise resource planning function, and a financial service module 143.

The financial server unit 2 includes a data server 4, a financial condition calculation server 5, a transaction server 6, and a position calculation server 7. The data server 4 is connected to the management terminal 3 and the point-of-sale terminal 1 via the communication network 100, and stores user data related to the point-of-sale terminal 1 and branch data corresponding to a plurality of branches. The user data includes product management data, transaction data, credit information, and point-of-sale terminal data. The credit information may include, for example, an asset record related to the asset status of a user who holds the point-of-sale terminal 1 and a credit record related to the user's credit status, and the user's public disclosure on Lianhe A credit transaction record of a public transaction record, the asset record, the credit record, and the public transaction record together constitute a credit record. In this embodiment, the credit record contains an indication that the user holds The total number of credit cards, at least one of the total number of credit cards, a revolving indicator indicating whether the user uses a cycle, and a total amount of credit related to the user. The transaction data may include, for example, a transaction record related to the user's transaction using the financial service module 143, and an interaction record related to the transaction status of the user with a bank corresponding to the financial server unit 2. In an embodiment, the interaction record includes a total number of products indicating that the user holds the product of the bank, a total number of interaction years indicating the number of years the user has been in contact with the bank, and an indication that the user has been in The total number of accounts of the bank's account opening, a total number of remittances indicating the number of remittances of the user in the bank in a section, a total remittance amount indicating the remittance amount of the user in the bank in the section, a At least one of the total number of remittance accounts indicating the number of remittance accounts of the user in the bank within the interval, and at least one of the value of the degree of contribution of the user's contribution to the bank in the interval. The value of the contribution degree can be obtained by the user's sum of the bank's fee and interest income generated in the interval. Wherein, after each period of time, the data server 4 stores the latest obtained credit information record and the interaction record.

The financial condition calculation server 5 is electrically connected to the data server 4 and is connected to the point-of-sale terminal 1 via the communication network 100. The transaction server 6 is electrically connected to the data server 4 and is connected to the point-of-sale terminal 1 via the communication network 100. The position calculation server 7 is electrically connected to the data server 4 and is connected to a plurality of branch-end devices (not shown) corresponding to the branches via an internal communication network 101 and to the sales via the communication network 100 Point terminal 1 to connect. In this embodiment, the financial server unit 2 is a server host with high computing power, and the data server 4, the financial condition calculation server 5, the transaction server 6, and the position calculation server 7 are respectively As a plurality of virtual machines in the servo host, the financial condition calculation server 5, the transaction server 6, and the position calculation server 7 are connected to the data server 4 through the internal communication network 101, respectively. However, in other embodiments of the present invention, the financial server unit 2 may also be a server unit composed of multiple separate servers, and the data server 4, the financial condition calculation server 5, and the transaction server The server 6 and the position calculation server 7 are a plurality of separate servers in the server unit.

It is worth mentioning that, because the sales checkout function provided by the point of sale terminal 1 can support payment methods such as credit cards, WeChat payments, and One-Cards, it is desirable to use the financial services provided by the present invention and hold the sales The user of the point terminal 1 opens an account corresponding to the point of sale terminal 1 at the bank corresponding to the financial server unit 2 so that the amount paid by consumers using payment methods such as credit card, WeChat payment, and card It can be deposited in an account corresponding to the point of sale terminal 1. In addition, when the user applies for financial services for the first time, the user can generate at least one financial service activation request to the data server 4 of the financial server unit 2 by operating the point-of-sale terminal 1 and transmitting through the communication network 100 When the data server 4 receives the at least one financial service activation request via the communication network 100, the data server 4 includes the at least one financial service activation request and the point of sale terminal location included in the point of sale terminal information. , And the branch location and branch identification information of each branch included in the branch information are transmitted to the location calculation server 7, when the location calculation server 7 receives the at least one financial service activation request, the location of the point of sale terminal, and When the branch locations are located, the location calculation server 7 obtains branch identification data corresponding to the branch location closest to the point of sale terminal location based on the point of sale terminal location and the branch locations, and via the internal communication network 101 Transmitting the at least one financial service activation request to a branch-end device corresponding to the branch identification information. When the branch-end device receives the at least one financial service activation request via the internal communication network 101, a branch service person holding the branch-end device can bring the branch-end device to visit the person holding the point-of-sale terminal 1 in person. The user uses the branch-end device to complete the online opening procedure of the financial service corresponding to the at least one financial service opening request. The at least one financial service opening request includes a deposit service opening request, a wealth management service opening request, and a loan. At least one of a service activation request and an insurance service activation request. It is worth noting that the opening of most financial services can be completed online. However, due to government regulations, the opening of a small number of financial services may require manual filling and signing of summer documents to complete the opening process. Regardless of whether the opening of the financial service is completed electronically or through the fractional operation of some electronic and manual operations, after the financial service is completed, the user can then use the point of sale terminal 1 to target A financial service request is made for the opened financial service.

Referring to FIG. 1, FIG. 2, and FIG. 3, the operations between the components of the point-of-sale terminal 1, the financial server unit 2, and the management terminal 3 will be described with reference to the embodiment of the financial service method of the present invention. This embodiment includes a procedure for storing product parameter data and a procedure for generating financial service conditions in this order. The product parameter data storage program includes steps 800 to 803, and the financial service condition generation program includes steps 807 to 815.

In step 800, the enterprise resource planning module 142 of the point-of-sale terminal 1 generates an input signal according to the input from the terminal input unit 11 and a user's input operation in a time interval, and correspondingly generates an input signal including Product management records for one product purchase record, one product sales record, and one product inventory record for a product. Among them, the product purchase record contains the purchase quantity of each product and its corresponding purchase time, as well as the purchase cost and payment terms of each product, and the product sales record contains the sales quantity of each product and its corresponding sales Time and payment conditions, and the sales amount of each product, the product inventory record contains the inventory quantity of each product. In this embodiment, the time interval may be set to one month, such as the 21st of each month to the 20th of the following month, but it is not limited thereto.

In step 801, the enterprise resource planning module 142 of the point-of-sale terminal 1 obtains a corresponding time interval according to the product management record, and contains at least one purchase parameter related to the purchase of the products, at least one related to the products The sales parameter related to the sales of the products, and the product parameter data of at least one inventory parameter related to the inventory of the products, and the product parameter data is transmitted to the management terminal 3 through the terminal communication unit 13. In this embodiment, the at least one purchase parameter includes at least one of an account payable turnover rate, an account payable turnover day, an average total purchase amount, and an item number of the purchase item, and the at least one sales parameter includes an application amount. Accounts receivable turnover rate, a accounts receivable turnover days, a net sales margin and an average total sales volume, the at least one inventory parameter includes a inventory turnover rate, a inventory turnover days, an average total inventory and At least one of a business cycle.

In step 802, when the management terminal 3 receives the product parameter data from the point-of-sale terminal 1 via the communication network 100, the management terminal 3 transmits the product parameter data to the financial server unit via the communication network 100 2 data server 4.

In step 803, when the data server 4 receives the product parameter data from the management terminal 3 via the communication network 100, the data server 4 stores the product parameter data into the product management data. Because the point-of-sale terminal 1 will transmit the product parameter data to the management terminal, so that the management terminal transmits the product parameter data to the financial server unit 2 for storage, and the data server 4 manages and stores the product parameter data This cloud-based management storage can centrally manage customer's product parameter data. If the customer has multiple branches, the point-of-sale terminal 1 set at each branch will send its corresponding product parameter data to the management terminal, so that The financial server unit 2 can store product parameter data corresponding to all branches of the customer. This mechanism is very different from the existing point-of-sale terminal that can only record product parameter data of one branch.

In step 807, the financial service module 143 of the point-of-sale terminal 1 generates at least one financial service request according to another input signal generated from the terminal input unit 11 and through the input operation of the user, and passes the The terminal communication unit 13 transmits the at least one financial service request to the transaction server 6 of the financial server unit 2.

In step 808, when the transaction server 6 receives the at least one financial service request from the point-of-sale terminal 1 via the communication network 100, the transaction server 6 transmits the at least one financial service request to the data server. 4.

In step 809, when the data server 4 receives the at least one financial service request, the data server 4 compares one of the stored product management data with the current time point (i.e., receives the at least one A time point for receiving a financial service request) The product parameter data of the shortest interval is used as the product parameter data of the current time interval, and the stored transaction data is obtained at the shortest interval from the current time point. The interaction record is used as an interaction record in the current time interval, and the stored credit information obtained at the shortest interval from the current time point is used as the credit record in the current time interval. The server 4 transmits the at least one financial service request, product parameter data of the current time interval, interaction records and credit records to the financial condition calculation server 5.

In step 810, when the financial condition calculation server 5 receives the at least one financial service request and product parameter data, interaction records, and credit records of the current time interval, the financial condition calculation server 5 uses a type of nerve The network algorithm is based on the at least one financial service request, product parameter data of the current time interval, interaction records and credit records, and the hidden state value calculated in the previous time interval and related to the hidden layer of a class of neural network models , Predicting a hidden state value related to the hidden layer of the neural network model in the current time interval, and calculating a credit rating related to the point of sale terminal based on the hidden state value in the current time interval.

It is worth noting that the hidden state value of the current time interval that is related to the hidden layer of the neural network model Based on the current time interval Product parameter data, interactive records and credit records The hidden state value of the previous time interval related to the hidden layer of this type of neural network model And the following formula (1). In addition, in this embodiment, By the current time interval Composed of product parameter data, interaction records, and credit records. However, in other embodiments of the present invention, Or just from the current time interval It is not limited to at least one of the product parameter data, interaction records and credit records. …(1)

among them, Is one of a linear rectification function and a leaky linear rectification function, In the current time interval And related to the weight matrix between the input layer and the hidden layer of this type of neural network model, Represents the previous time interval The hidden layer of this type of neural network model is passed to the current time interval The weight matrix of the hidden layer of this type of neural network model. It is worth noting that when the current time is the first time interval (that is, in the initial state), at this time, the previous time interval (the 0th time interval) is related to the neural network model The hidden state value of the hidden layer is set to a preset initial value. Even if the hidden state value predicted by the financial condition calculation server 5 for the hidden layer of the neural network model for the first time interval is inaccurate, the embodiment of the financial service method of the present invention may still be based on Subsequent product parameter data received gradually converge to a better prediction model to predict more and more accurate hidden state values. In addition, this type of neural network algorithm is different for different training data, and the type of neural network model that is converged will also be different. With this autonomous learning characteristic of this type of neural network algorithm, the present invention uses neural network-like The roadshow algorithm is used to estimate the credit rating. It can build a prediction model suitable for assessing its credit rating according to the small and small profit enterprises of different industry categories.

Credit rating of the point of sale terminal Is based on the hidden state value of the hidden layer of the neural network model according to the current time interval And the following formula (2). …(2)

among them, Is a normalized exponential function, In the current time interval And it is related to the weight matrix between the hidden layer and the output layer of this type of neural network model.

In step 811, the financial condition calculation server 5 generates according to the credit rating and transmits at least one pair of financial service conditions corresponding to at least one financial service request to the point of sale terminal 1 via the communication network 100.

In this embodiment, the at least one financial service request includes at least one of a deposit service request, a wealth management service request, a loan service request, and an insurance service request. The at least one financial service condition includes a deposit service condition, a At least one of a financial service condition, a loan service condition, and an insurance service condition. The deposit service condition includes, for example, at least one deposit plan, and each deposit plan includes a corresponding deposit interest rate. The financial service condition includes, for example, at least one investment portfolio solution. The loan service condition includes, for example, at least one loan scheme, and each loan scheme includes a corresponding loan amount and a loan interest rate. The insurance service condition includes, for example, at least one insurance scheme.

For example, in terms of loan service conditions, when the credit rating is higher, the loan amount corresponding to the loan service conditions will be higher and the loan interest rate will be lower.

In step 812, when the financial service module 143 of the point-of-sale terminal 1 receives the at least one financial service condition from the financial servo order via the terminal communication unit 13, the financial service module 143 changes the at least one financial service The service conditions are presented on the terminal display unit 12.

In step 813, the financial service module 143 of the point-of-sale terminal 1 also generates another input signal corresponding to the at least one finance according to another input signal generated from the terminal input unit 11 and through the input operation of the user. The transaction request of one of the service conditions is transmitted to the transaction server 6 of the financial server unit 2 via the terminal communication unit 13. By integrating financial functions into the point-of-sale terminal 1, the point-of-sale terminal 1 itself contains modules for financial services, and customers can directly see the individualized financials of each financial service on the point-of-sale terminal 1. Records of services and account transactions, interest, and investment gains and losses.

In step 814, when the transaction server 6 receives the transaction request from the point of sale terminal 1 via the communication network 100, the transaction server 6 generates a transaction record corresponding to the transaction request according to the transaction request, and The transaction record is transmitted to the data server 4.

In step 815, when the data server 4 receives the transaction record, the data server 4 stores the transaction record into the transaction data.

In summary, in the financial service method of the present invention, the product parameter data is transmitted to the management terminal 3 by the point of sale terminal 1, so that the management terminal 3 transmits the product parameter data to the financial server unit 2, thereby making The financial servo unit 2 stores the product parameter data into the product management data, and the financial servo unit 2 according to the product parameter data, interaction records and credit records in the current time interval, and the hidden data calculated in the previous time interval. The credit rating predicted by the hidden status value of the layer is generated according to the customer's operating status and transaction status, whereby the financial server unit 2 can provide customers with customers based on the at least one financial service condition generated by the credit rating. Institutionalized financial services can also reduce the risks of banks from providing financial services such as lending, so it can indeed achieve the purpose of cost invention.

However, the above are only examples of the present invention. When the scope of implementation of the present invention cannot be limited by this, any simple equivalent changes and modifications made according to the scope of the patent application and the contents of the patent specification of the present invention are still Within the scope of the invention patent.

1‧‧‧point of sale terminal

11‧‧‧Terminal input unit

12‧‧‧ terminal display unit

13‧‧‧Terminal communication unit

14‧‧‧Terminal Processing Unit

141‧‧‧Sales Management Module

142‧‧‧Enterprise Resource Planning Module

143‧‧‧Financial Services Module

2‧‧‧Financial Servo Unit

3‧‧‧ Management

4‧‧‧ Data Server

5‧‧‧Financial condition calculation server

6‧‧‧Transaction Server

7‧‧‧Position calculation server

100‧‧‧Communication Network

101‧‧‧ Intercom Network

800 ~ 803‧‧‧step

807 ~ 815‧‧‧step

Other features and effects of the present invention will be clearly presented in the embodiment with reference to the drawings, in which: FIG. 1 is a block diagram illustrating an embodiment of a point-of-sale terminal of the present invention and an embodiment of a financial servo unit of the present invention, And a management terminal connected to the point-of-sale terminal and the financial server unit via the communication network; FIG. 2 is a flowchart illustrating a product parameter data storage procedure in the embodiment of the financial service method of the present invention; and FIG. 3 is A flowchart illustrating a procedure for generating a financial service condition in the embodiment of the financial service method of the present invention.

Claims (10)

A financial service method is implemented by a point-of-sale terminal, which can provide an enterprise resource planning function, and is connected to a management terminal and a financial server unit through a communication network, and the management terminal also passes the communication network Connected to the financial server unit, the financial server unit stores user data related to the point of sale terminal, the user data includes product management data, and the financial service method includes the following steps: (A) The point of sale terminal is in a The time interval generates a product management record including a product purchase record, a product sales record, and a product inventory record related to multiple products through a user's input operation; (B) the point of sale terminal according to the product management record, Obtain a corresponding time interval that includes at least one purchase parameter related to the purchase of these products, at least one sales parameter related to the sale of these products, and at least one inventory parameter related to the inventory of the products Product parameter data, and transmits the product parameter data to the management terminal, so that the management terminal transmits the product parameter data The data is transmitted to the financial server unit; (C) The point of sale terminal generates and transmits at least one financial service request to the financial server unit through the user's input operation, so that the financial server unit will store the product management data The product parameter data of the time interval with the shortest interval between one of the current time points and the current time interval is used as the product parameter data of the current time interval, and a type of neural network algorithm is used to at least based on the product parameter data of the current time interval and the previous time interval. Calculate the hidden state value of the hidden layer of a class of neural network models, predict the hidden state value of the hidden layer of the current time interval, and predict the hidden state value of the current time interval. Calculate a credit rating related to the point of sale terminal, and generate at least one pair of financial service conditions that should correspond to at least one financial service request based on the credit rating; and (D) when the point of sale terminal receives the financial service When the at least one financial service condition generated by the servo unit, the point of sale terminal presents the at least one gold Financial service conditions. The financial service method according to claim 1, wherein, in step (C), the hidden state value of the current time interval related to the hidden layer of the neural network model Also obtained according to the following formula: among them, Is one of a linear rectification function and a leaky linear rectification function, In the current time interval And related to the weight matrix between the input layer and the hidden layer of this type of neural network model, In the current time interval Product parameter data, Represents the previous time interval The hidden layer of this type of neural network model is passed to the current time interval The weight matrix of the hidden layer of this type of neural network model, The hidden state value of the hidden layer related to the neural network model of the previous time interval. The financial service method according to claim 1, wherein in the step (C), the credit rating of the point-of-sale terminal Also obtained according to the following formula: among them, Is a normalized exponential function, In the current time interval And related to the weight matrix between the hidden layer and the output layer of this type of neural network model, The hidden state value corresponding to the hidden layer of the neural network model representing the current time interval. The financial service method according to claim 1, wherein in the step (B), the at least one purchase parameter includes a accounts payable turnover rate, a accounts payable turnover day, an average total purchase amount, and a purchase item The at least one inventory parameter includes at least one of a accounts receivable turnover rate, a accounts receivable turnover day, a net sales margin, and an average total sales amount. The at least one inventory parameter includes At least one of an inventory turnover rate, an inventory turnover day, an average inventory total, and an operating cycle. According to the financial service method described in claim 1, the user data further includes transaction data, and after step (D), further includes the following steps: (E) The point-of-sale terminal generates an association with A transaction request for one of the at least one financial service condition, and the transaction request is transmitted to the financial server unit, so that the financial server unit stores a transaction record corresponding to the transaction request in the transaction data. A financial service method is implemented by a financial server unit, which is connected to a management terminal and a point-of-sale terminal via a communication network. The point-of-sale terminal can provide an enterprise resource planning function, and the financial server unit stores There is user data related to the point of sale terminal, the user data includes product management data, and the financial service method includes the following steps: (A) When the financial server unit receives a corresponding A time interval that includes at least one purchase parameter related to the purchase of multiple products, at least one sales parameter related to the sale of these products, and at least one inventory parameter related to the inventory of the products In the case of data, the financial server unit stores the product parameter data into the product management data; (B) When the financial server unit receives at least one financial service request, the financial server unit stores the product management data in the product management data The product parameter data of the time interval with the shortest interval between the current time point and the current time interval is used as the product of the current time interval. Parameter data, and use a class of neural network algorithms to predict at least the current time interval based on the product parameter data of the current time interval and the hidden state value of the hidden layer of a type of neural network model. A hidden state value related to the hidden layer of the neural network model, and a credit rating related to the point-of-sale terminal is calculated based on the hidden state value of the current time interval; and (C) the financial server unit is based on The credit rating generates at least one pair of financial service conditions that should correspond to at least one financial service request, and transmits the at least one financial service condition to the point of sale terminal. The financial service method according to claim 6, wherein, in step (B), the hidden state value of the current time interval related to the hidden layer of the neural network model Also obtained according to the following formula: among them, Is one of a linear rectification function and a leaky linear rectification function, In the current time interval And related to the weight matrix between the input layer and the hidden layer of this type of neural network model, In the current time interval Product parameter data, Represents the previous time interval The hidden layer of this type of neural network model is passed to the current time interval The weight matrix of the hidden layer of this type of neural network model, The hidden state value of the hidden layer related to the neural network model of the previous time interval. The financial service method according to claim 6, wherein, in step (B), the credit rating of the point of sale terminal Also obtained according to the following formula: among them, Is a normalized exponential function, In the current time interval And related to the weight matrix between the hidden layer and the output layer of this type of neural network model, The hidden state value corresponding to the hidden layer of the neural network model representing the current time interval. The financial service method according to claim 6, wherein in the step (A), the at least one purchase parameter includes an account payable turnover rate, an account payable turnover day, an average total purchase amount per month, and one purchase item At least one of the items, the at least one sales parameter includes at least one of an account receivable turnover rate, an account receivable turnover day, a net sales margin and an average monthly total sales amount, the at least one inventory The parameters include at least one of an inventory turnover rate, an inventory turnover day, an average monthly total inventory, and an operating cycle. According to the financial service method described in claim 6, the user data further includes transaction data, and after step (C), further includes the following steps: (D) when the financial server unit receives a related to the at least one financial service When one of the conditions is a transaction request, the financial server unit stores a transaction record corresponding to the transaction request into the transaction data.