WO2018228408A1

WO2018228408A1 - Method and apparatus for settling transaction, server, and storage medium

Info

Publication number: WO2018228408A1
Application number: PCT/CN2018/090977
Authority: WO
Inventors: 张帅
Original assignee: 腾讯科技（深圳）有限公司
Priority date: 2017-06-16
Filing date: 2018-06-13
Publication date: 2018-12-20
Also published as: CN109146440A; CN109146440B

Abstract

Disclosed are a method and apparatus for settling a transaction, a server, and a storage medium. The method comprises: acquiring a transaction record to be settled; using a rule matching model obtained by means of pre-training, to determine the matching degree of the transaction record respectively with pre-set multiple transaction settlement rules; according to the matching degree of the transaction record respectively with the pre-set multiple transaction settlement rules, determining, from the pre-set multiple transaction settlement rules, a target transaction settlement rule required for a transaction settlement of the transaction record; and based on the target transaction settlement rule, carrying out transaction settlement on the transaction record. The solution of the present application can reduce the cases in which a transaction settlement fails due to the inability to match a transaction settlement rule, thereby reducing a delay in settlement and entry of a payee in a payment platform.

Description

Transaction settlement method, device, server, and storage medium

This application claims the priority of the Chinese Patent Application, filed on Jun. 16, 2017, the number of which is hereby incorporated by reference. In this application.

Technical field

The present application relates to the field of transaction data processing technologies, and in particular, to a transaction settlement method, apparatus, server, and storage medium.

Background technique

With the development of network technology, the application of third-party payment and other payment platforms has become increasingly widespread.

After the transaction payment is completed based on the payment platform (eg, transfer, merchandise payment, cash withdrawal, etc.), the payment platform generates a transaction record, and the transaction record generally records the parties involved in the transaction, the payment method, the transaction platform, and the payment amount. And a variety of transaction information. The payment platform performs transaction settlement (also called payment settlement) on the transaction record according to the transaction charging rule matched with the transaction record. For example, the payment platform determines the required deduction according to the charging rule matched with the transaction record. After the fee or service fee is deducted from the payment amount, the remaining payment amount is deposited into the designated account of the payee in the transaction to complete the transaction settlement.

Among them, the transaction settlement rules are not only related to the payee in the transaction, but also related to one or more of various factors such as payment method, trading platform and payment amount. However, factors such as the payment method, trading platform and payment amount of the payee of the transaction often change, and once these factors change, there may be a transaction record in the payment platform that is not configured with the payee of the transaction. The matching transaction settlement rules result in the inability to complete the settlement of the transaction, which in turn leads to delays in the settlement of the payee's settlement.

Summary of the invention

In view of this, the present application provides a transaction settlement method, apparatus, server, and storage medium to reduce the occurrence of transaction settlement failure due to failure to match the transaction settlement rule, thereby reducing the payment settlement delay of the payee.

To achieve the above object, the present application provides a transaction settlement method, including:

Obtain transaction records to be settled;

Determining, by the pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, wherein the rule matching model is based on a plurality of historical transaction records before the current time and the history The trading settlement rules corresponding to the transaction records are trained;

Determining a target transaction settlement rule required for transaction settlement of the transaction record from the preset plurality of transaction settlement rules according to a matching degree of the transaction record with a preset plurality of transaction settlement rules;

Transaction settlement is performed on the transaction record based on the target transaction settlement rule.

The application provides a transaction settlement device, including:

a record acquisition unit for obtaining a transaction record to be settled;

a rule matching unit, configured to determine, by using a pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, wherein the rule matching model is based on multiple history before the current time Transaction records and training settlement rules corresponding to the historical transaction records;

a rule determining unit, configured to determine, according to a matching degree of the transaction record and a preset plurality of transaction settlement rules, from the preset plurality of transaction settlement rules, determining a transaction settlement required for the transaction record Target transaction settlement rules;

a transaction settlement unit, configured to perform transaction settlement on the transaction record based on the target transaction settlement rule.

The application provides a server, including:

Processor and memory;

Wherein the processor is configured to execute a program stored in the memory;

The memory is for storing a program, the program being at least for:

Obtain transaction records to be settled;

The present application also provides a storage medium having stored therein computer executable instructions for performing a transaction settlement method as described above.

The present application also provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform a transaction settlement method as described above.

According to the above technical solution, in the present application, after obtaining the transaction record to be processed, the rule matching model obtained by the pre-training may be used to determine the matching degree of the transaction record to the preset multiple transaction settlement rules respectively. And according to the matching degree of the transaction record with the preset multiple transaction settlement rules, from the plurality of transaction settlement rules, a transaction settlement rule is selected to perform transaction settlement on the transaction record, thereby reducing the occurrence of the payment platform. The transaction settlement rules are missing, which makes it impossible to settle the transaction records, which reduces the delay in transaction settlement.

DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only examples of the present application. One of ordinary skill in the art can also obtain other drawings based on the drawings provided without the inventive effort.

FIG. 1 is a schematic diagram showing a composition structure of an application scenario to which a transaction settlement method of the present application is applied;

2 is a schematic diagram showing the structure of a server to which a transaction settlement method of the present application is applied;

3 is a schematic flow chart showing an embodiment of a transaction settlement method of the present application;

4 is a schematic diagram showing input parameters and output results of an RNN model of the present invention as a rule conversion model;

FIG. 5 is a schematic diagram showing a training process of training an RNN model as a rule conversion model according to the present application; FIG.

6 is a schematic flow chart of a transaction settlement method of the present application in an application scenario;

FIG. 7 is a schematic diagram showing an execution flow of a transaction settlement method of the present application; FIG.

FIG. 8 is a block diagram showing the structure of an embodiment of a transaction settlement device of the present application.

detailed description

FIG. 1 is a schematic structural diagram of an application scenario in which the solution of the present application is applied, where the application scenario includes:

The payment platform 101, the transaction platform 102, and the user's terminal 103.

The payment platform 101 can include at least one payment server 1011. For example, the payment platform may be a third-party payment platform, or may be a platform for a payment system such as a bank that can realize transaction payment by both parties. The trading platform 102 is a service platform that can complete product transactions between the provider and the user. The trading platform 102 can include at least one trading server 1021.

The merchant in the trading platform 102 is a merchant registered with the payment platform 101, or a merchant having a contractual relationship with the payment platform 101.

Accordingly, the user is a registered user on the payment platform. The user can access the trading platform 102 through the terminal 103 to conduct product transactions with the merchants in the trading platform, and after the product transaction is reached, send a payment request to the payment platform through the terminal 103, the payment request carrying the identifier of the paying party (ie, The user's identification), the identity of the payee (ie, the identity of the merchant), the transaction amount, the payment channel (eg, UnionPay card payment, credit card payment, account balance payment, etc.). The payment request may also carry: an identification of the trading platform, a product type of the transaction, and the like.

The payment platform generates a transaction record according to the payment request submitted by the user, and the transaction record may include the identifier of the above payment party carried by the payment request, the identifier of the payee, the payment channel, the identifier of the transaction platform, and the product type of the transaction. The payment platform may also store the access channel and the settlement mode as information in the transaction record according to the identifier of the payee, or the access channel and the settlement mode of the payee accessing the payment platform. The access channel indicates a specific manner used by the payee to access the payment platform. For example, the access channel may include: accessing the payment platform through a mobile terminal or a wired network terminal, and the settlement mode indicates the type of the transaction, for example, The settlement mode can include: transaction payment, transfer, and withdrawal.

The payment platform may set the frequency of transaction settlement according to needs, for example, one transaction settlement is performed every natural day, or the transaction settlement may be performed every specified time. Of course, the payment platform may also be different for different trading platforms or different. The merchant sets the frequency of different transaction settlements.

When the current settlement condition of the transaction is satisfied, the payment platform acquires at least one transaction record currently to be settled; for each transaction record, determines a transaction settlement rule matching the transaction record, and performs the transaction record according to the transaction settlement rule Transaction settlement. For example, the transaction settlement rule may be a charging rule for determining a service fee or a fee for deduction of the transaction record, and the deduction amount required for the payment settlement of the transaction record may be determined according to the charging rule. And deducting the deduction amount from the transaction amount of the transaction record, and storing the remaining transaction amount in the designated account of the payee corresponding to the transaction record.

It can be understood that, in the application scenario shown in FIG. 1 , the transaction platform included in the scenario is taken as an example, in which the transaction parties complete the transaction payment based on the payment platform after generating the product transaction. However, it can be understood that the application scenario to which the solution of the present application is applied is not limited to that shown in FIG. 1. In the actual application, the user pays according to the payment platform to the merchant, and the user withdraws the cash based on the payment platform. The platform generates transaction records and settles transactions for transaction records when transaction settlement is required.

The inventor of the present application studies the transaction settlement process of the payment platform and finds that the transaction settlement rules are related to the payment channels set by the merchants, the merchants, the transaction platform to which the merchant belongs, the settlement mode set by the merchant, and the like. When the merchant changes one or more factors, if the management personnel of the payment platform fail to enter the corresponding transaction settlement rules in time, or establish the correspondence between the transaction settlement rules and the changed factors in time, it may result in The transaction settlement rules are missing, and the transaction settlement rules corresponding to the transaction records are not configured in the payment platform.

For example, suppose a merchant's payment channel includes credit card payment and debit card payment. If the merchant adds a UnionPay card to quickly pay for this payment channel, the payment platform does not correspond to the merchant and the UnionPay card quick payment. In the transaction settlement rule, if there is a transaction record for the merchant and the UnionPay card is used for fast payment, the transaction settlement rule corresponding to the transaction record cannot be matched, and the transaction settlement fails.

However, in the process of settlement of the transaction by the payment platform for the transaction record, if the transaction settlement rule matching the transaction record cannot be queried, the transaction record is marked as the transaction record of the settlement abnormality, and the alarm of the settlement abnormality is output. Prompt, and wait for the professional's intervention confirmation and problem repair, resulting in the inability to complete the transaction settlement in time, making the transaction settlement delay. At the same time, in the case of a settlement abnormality, there is also a high requirement for the business knowledge reserve of the person handling the settlement abnormality, resulting in a high complexity of the exception handling.

In order to solve the billing abnormality, reduce the delay of the payment settlement, and reduce the complexity of the billing abnormality processing when the settlement abnormality occurs, the embodiment of the present application provides a transaction settlement method, and the transaction settlement method can be applied to payment. The server of the platform to settle the payment records.

For example, referring to FIG. 2, a schematic structural diagram of a server of a payment platform to which the transaction settlement method of the embodiment of the present application is applied is shown.

In FIG. 2, the server 200 can include a processor 201, a memory 202, a communication interface 203, an input unit 204, a display 205, and a communication bus 206.

The processor 201, the memory 202, the communication interface 203, the input unit 204, and the display 205 all communicate with each other via the communication bus 206.

In the embodiment of the present application, the processor 201 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a digital signal processor (DSP), an application specific integrated circuit. (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic devices.

The processor can call a program stored in the memory 202.

The memory 202 is configured to store one or more programs, and the program may include program code, where the program code includes computer operation instructions. In the embodiment of the present application, at least the program for implementing the following functions is stored in the memory:

Obtain transaction records to be settled;

The memory 202 may include a high speed random access memory, and may also include a nonvolatile memory such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

The communication interface 203 can be an interface of a communication module, such as an interface of a GSM module.

The server of the present application may further include an input unit 204, which may include an input device such as a keyboard, a mouse, or the like.

The display 205 includes a display panel or the like.

Of course, the server structure shown in FIG. 2 does not constitute a limitation on the server in the embodiment of the present application. In practical applications, the server may include more or less components than those shown in FIG. 2, or combine some components.

The transaction settlement method of the embodiment of the present application will be described in detail below in conjunction with the above commonalities of the present application.

Referring to FIG. 3, it is a schematic flowchart of an embodiment of a transaction settlement method according to the present application. The method in this embodiment may be applied to a server of a payment platform, and the method may include:

S301. Acquire a transaction record to be settled.

Wherein, the transaction record includes transaction information of at least one attribute.

Among them, different attributes characterize the transaction records from different dimensions. For example, the attributes (or dimensions) that may be involved in the transaction record may include: the payee of the transaction record, the transaction amount of the transaction record, the payment method of the transaction record, The trading platform of the transaction record, the payment channel of the transaction (also known as payment method, such as credit card payment, UnionPay card payment, etc.), the product type of the transaction, the access channel of the payee accessing the payment platform, and the settlement mode (eg Trade payment, transfer or withdrawal) and other attributes.

In this embodiment of the present application, the transaction information may include transaction information of one or more of the above various attributes.

The transaction information of each attribute can be considered as the specific value of the transaction record under this attribute. For example, if the transaction record includes the transaction information under the attribute of the payee, the transaction record 1 records the transaction information of the payee A, and the transaction record 2 may record the payee B. Transaction information.

S302. Detect, according to the transaction information of the at least one attribute in the transaction record, whether the transaction settlement rule corresponding to the transaction record is configured in the payment platform.

For example, the payment platform is preset with a plurality of different transaction settlement rules, and each transaction settlement rule corresponds to at least one transaction information combination, the transaction information combination includes at least one transaction information, and the different transaction information combinations are included. The type and amount of transaction information can vary. Correspondingly, according to the correspondence between the transaction settlement rule and the transaction information combination, it is possible to detect whether there is a transaction settlement rule matching the transaction information of at least one attribute in the transaction record.

For example, the correspondence between the transaction settlement rules and the transaction information combination may include:

When the transaction information is combined into the payee A, and the payment channel is a credit card payment, the transaction settlement rule is: the deduction amount required to be deducted is equal to the payment amount multiplied by the first billing rate (eg, the first billing fee) Rate is 0.6%);

When the transaction information is combined into the payee A, the payment amount is greater than M2, and the payment channel is the fast payment of the UnionPay card, the transaction settlement rule is: the deduction amount required to be deducted is the preset amount (for example, 1,000 yuan).

When the transaction information is combined into the payee B, and the payment channel is a UnionPay card payment, the transaction settlement rule is: the deduction amount required to be deducted is equal to the payment amount multiplied by the second charging rate (eg, the first billing rate) The rate is 1.5%).

It can be understood that, in practical applications, the payment platform can also set which transaction information in the transaction record belongs to the key transaction information required to match the transaction settlement rule, and therefore, if there is a transaction information combination and a few included in the transaction record If the key transaction information is completely matched, the transaction settlement rule corresponding to the transaction information combination may be determined as the transaction settlement rule corresponding to the transaction record.

It can be understood that the above is only an example of detecting whether a transaction settlement rule corresponding to the transaction record is stored. In an actual application, it is also possible to detect whether there is a corresponding transaction record by other means. The transaction settlement rules are not limited here.

It should be noted that due to various factors such as the change of the relevant attributes of the signing merchants in the payment platform, the payment platform may not be configured with the corresponding transaction settlement rules or the transaction settlement rules are missing in time when the transaction records are generated. There is currently no transaction settlement rule configured for the transaction record in the payment platform. For the above situation, the embodiment of the present application can process the rule matching model, refer to S303.

S303, when there is no transaction settlement rule corresponding to the transaction record in the payment platform, using a rule matching model obtained by the pre-training, determining a matching degree between the transaction record and the preset multiple transaction settlement rules;

Optionally, after the rule matching model is established, the rule matching model needs to be trained. In the process of training the rule matching model, a matching rule for determining the matching degree may be mined, and the matching rule may be: The degree of influence of the different transaction information contained in the attribute, and the matching of the transaction record with the preset multiple transaction settlement rules. For example, when the attribute is the payee, the payee can have multiple possibilities, and different payees have different influences on the transaction record containing the payee and the matching of different transaction settlement rules. Correspondingly, the matching rule in the model and the transaction information of the at least one attribute in the transaction record may be matched according to the rule, and the rule matching model is used to calculate the transaction record and the preset multiple transaction settlement rules respectively. The degree of matching between.

For example, in a possible implementation manner, each transaction includes different transaction information, and the degree of influence on the transaction record matching the preset multiple transaction settlement rules may be: each transaction contains different transaction information respectively. The matching scores matched by the preset multiple transaction settlement rules, and the different transaction information contained in each attribute affect the transaction records respectively with the preset multiple transaction settlement rules. In this way, after obtaining the transaction information of the at least one attribute included in the transaction record, respectively, for each of the preset transaction settlement rules, respectively determine the matching scores of each transaction information included in the transaction record and the transaction settlement rules, and each transaction. The information affects the influence weight of the transaction record matching the transaction settlement rule, and calculates a weighted sum of the matching scores corresponding to the transaction information of the at least one attribute included in the transaction record, thereby determining the calculated score as the transaction record The degree of matching with the transaction settlement rules.

It can be understood that the matching degree of the transaction record and the transaction settlement rule can be characterized by the matching probability. In this case, the matching score referred to above may be a value greater than or equal to zero and less than 1. However, it can be understood that the degree of matching by matching probability is only an implementation. In practical applications, the degree of matching between transaction records and transaction settlement rules can also be scored (for example, the score can be from 0 to 100). Any numerical representation), wherein the higher the score, the higher the match between the transaction record and the transaction settlement rule. Of course, the degree of matching between the transaction record and the transaction settlement rule can also be represented by the matching level between the transaction record and the transaction settlement rule, which is not limited herein.

It should be noted that, here, for the sake of easy understanding, an example of a matching rule is introduced as an example, but in practical applications, according to the difference of the network model adopted by the rule matching model, the rule matches the model. The specific implementation of the mining rules that are mined may be more complicated.

In the embodiment of the present application, the rule matching model is trained by using various transaction information included in a plurality of historical transaction records before the current time. Among them, in order to facilitate the distinction, the transaction record before the current time is referred to as a historical transaction record.

Among them, the rule matching model obtained by training may have various possible situations. For example, the rule matching model may be a Recurrent Neural Networks (RNN) model. Of course, the rule matching model can also be obtained through other algorithms or model training, for example, based on Bayesian classifier model, Support Vector Machine (SVM) model, symbolic reasoning algorithm, and genetic algorithm based competition for elimination. The rule matching model is trained by one or more of different algorithms or models, such as a deep learning algorithm and a heterogeneous learner, and is not limited herein.

It should be noted that, in this embodiment, after the transaction record is acquired, according to the transaction information of at least one attribute in the transaction record, it is detected whether a transaction settlement rule corresponding to the transaction record is configured in the payment platform, and When the payment platform does not have a transaction settlement rule corresponding to the transaction record, the rule matching model obtained by the pre-training is used to determine the matching probability between the transaction record and the different transaction settlement rules as an example. However, it can be understood that, in practical applications, even if the transaction settlement rule corresponding to the transaction record is configured in the payment platform, after obtaining the transaction record, the rule matching model obtained by the pre-training can be used to determine the transaction. Record the probability of matching with different transaction settlement rules, in order to finally select a more suitable transaction settlement rule to settle the transaction record. Therefore, after the transaction record is acquired, the operation of step S302 may not be performed, and the rule matching model may be directly used to determine the matching probability between the transaction record and the different transaction settlement rules.

S304. Determine, according to the matching degree between the transaction record and the preset plurality of different transaction settlement rules, determine, from the preset plurality of different transaction settlement rules, the target transaction settlement with the highest degree of matching with the transaction record. rule.

In order to facilitate the distinction, the transaction matching rule with the highest matching degree is called the target transaction settlement rule. Wherein, the target transaction settlement rule can be used as a transaction settlement rule required for the transaction record of the transaction record.

For example, the matching probability of the transaction settlement rule A matching the transaction record is 0.6, and the matching probability of the transaction settlement rule B matching the transaction record is 0.3, and the matching probability of the transaction settlement rule C matching the transaction record is 0.1, then The transaction settlement rule A with the highest matching probability is taken as the target transaction settlement rule.

It can be understood that the present embodiment illustrates the matching degree between the transaction record and the transaction settlement rule by using the matching probability as an example, but it can be understood that when the transaction record and the transaction settlement rule are represented by the score or the rank. When the matching degree is the same, the transaction settlement rule with the highest score or the highest matching level can be selected as the target transaction settlement rule.

Of course, selecting the target transaction settlement rule with the highest matching degree is only a preferred method. In the actual application, a transaction settlement rule may be selected as the target transaction settlement from the transaction settlement rule whose matching probability or matching score exceeds the preset threshold. Rules, no restrictions here.

S305, the transaction settlement is performed on the transaction record according to the target transaction settlement rule.

After the target transaction settlement rule is determined, according to the transaction settlement rule, there may be multiple ways of settlement of the transaction record, which may be different according to different transaction settlement rules and application scenarios of transaction settlement. No restrictions.

It can be seen that, in the embodiment of the present application, after acquiring the transaction record to be processed, the rule matching model obtained by the pre-training may be utilized, and at least two transaction information based on the transaction information of the at least one attribute included in the transaction record, Calculating a matching probability that the transaction record matches the preset multiple transaction settlement rules respectively, and using the target transaction settlement rules with the highest matching degree with the transaction record in the plurality of transaction settlement rules, the settlement of the transaction record, This reduces the delay in transaction settlement due to the inability to complete the transaction settlement of the transaction record.

S306. Generate a transaction estimation voucher for the transaction record, and store the transaction estimation voucher.

The transaction estimation voucher may include: transaction information of at least one attribute included in the transaction record, a target transaction settlement rule corresponding to the transaction record, and a settlement amount for transaction settlement of the transaction record. The settlement amount of the transaction record can be determined by the transaction settlement of the transaction record, and the amount determined to be transferred to the payee corresponding to the transaction record. In order to determine the payee of the transaction, in the case where a transaction estimate voucher needs to be generated, the transaction record includes at least the payee of the transaction.

S307. When receiving the deviation compensation amount input by the user for the transaction estimation voucher, storing a correspondence between the deviation compensation amount and the payee of the transaction record, so as to receive the payment according to the deviation compensation amount. The party will compensate for the settlement amount of income after the current time.

For example, the amount of the settlement of the most recent income after the current time can be compensated according to the deviation compensation amount.

The purpose of generating the transaction estimation voucher is that after completing the transaction settlement, the user (eg, the service maintenance personnel of the payment platform, etc.) can analyze the target transaction settlement rule corresponding to the transaction record and settle according to the transaction estimation voucher. The amount is correct. If there is a problem with the settlement amount due to an error in the target transaction settlement rule, the user re-configures the transaction with at least one attribute of the transaction record in the payment platform after analyzing the transaction settlement rule that actually matches the transaction record. The transaction settlement rule of the information matching, and the payment platform or the user recalculates the actual settlement amount corresponding to the transaction record through the terminal. For the sake of distinction, the settlement amount actually required for settlement of the transaction record is referred to as the actual settlement amount.

Correspondingly, the amount of deviation between the settlement amount and the settlement amount can be calculated based on the actual settlement amount and the target transaction settlement rule, and the deviation compensation amount can be obtained. For example, the deviation compensation amount may be: calculating a difference between the settlement amount and the actual settlement amount by using a target transaction settlement rule.

After the user determines the offset compensation amount, the deviation compensation amount may be input or input into the server of the payment platform through the terminal. In this way, if the server acquires the transaction record corresponding to the payee again after the current time, then the transaction settlement of the transaction record is completed, and the account corresponding to the payee is transferred to the transaction settlement. Before the settlement amount, the sum of the deviation compensation amount and the settlement amount may be used as the amount to be transferred to the designated account.

For example, if the deviation compensation amount obtained by subtracting the settlement amount from the actual settlement amount by the target transaction settlement rule is -20 yuan, the amount of 20 yuan deducted from the settlement amount of the income required by the payee after the current time may be The settlement amount after deducting 20 yuan will be deposited into the account designated by the payee.

It can be understood that the target transaction settlement rule corresponding to the transaction record is determined by the rule matching model obtained through the training, and the settlement amount required for the transaction record is calculated by using the target transaction settlement rule, and the transaction of the transaction record is completed. After the settlement, the manner of adjusting the settlement amount of the transaction settlement required for the transaction record may be various, and is not limited to the implementation manners of step S306 and step S307.

It should be noted that the steps S307 and S307 are optional steps, and the purpose thereof is only to improve the accuracy of transaction settlement for the transaction record, and if only the delay of the transaction settlement is reduced, the step S306 and the step S307 may not be performed.

In addition, in S302 of the embodiment, if it is detected that there is a transaction settlement rule corresponding to the transaction record in the payment platform, the transaction settlement settlement corresponding to the transaction record may be directly calculated based on the transaction settlement rule corresponding to the transaction record. Amount, and complete the transaction settlement.

In order to facilitate the process of understanding the training rule matching model, the RNN model obtained by training the rule matching model is taken as an example. For example, referring to FIG. 4, it shows a schematic diagram of input parameters and output results in the process of training the RNN model in the present application. .

In FIG. 4, the parameters of the RNN model to be trained during the training process include: a plurality of transaction information included in the historical transaction record, such as a merchant (or a payee), a trading platform, a product type of the transaction, Payment method, access channel, settlement mode, payment amount, etc., and the result of the RNN model output is: the probability of matching between the historical transaction record and a plurality of different transaction settlement rules. For example, the right side of the RNN model in Figure 4 represents a variety of transaction settlement rules. For example, 0.6% means that the transaction settlement rule is: the deduction amount is equal to the product of the transaction amount and the billing rate of 0.6%; for example, 1 yuan per pen. Indicates that the transaction settlement rule is: the deduction amount corresponding to each transaction record is 1 yuan; for example, the package amount is 10,000, indicating that the transaction settlement rule is: within a specified time period (such as one year or one month), if the deduction has been deducted If the payee of the transaction record has a fixed deduction amount of 10,000 yuan, there is no need to deduct the deduction amount from the transaction record. For different transaction settlement rules, the matching probability corresponding to each transaction settlement rule can be output through the RNN model.

The process of the training rule matching model of the present application is introduced by taking the rule matching model as the RNN model as an example. For example, referring to FIG. 5, the training of the present application is used to determine transaction records and different transaction settlement rules. Schematic diagram of the training process of the RNN model of the matching probability, the process may include:

S501: Acquire a plurality of historical transaction records required for training and actual transaction settlement rules that each historical transaction record actually matches.

Wherein each historical transaction record includes historical transaction information of at least one attribute.

Among them, each historical transaction record is used as a training sample in the training process, and the historical transaction information in the historical transaction record is the training data of the training sample.

The historical transaction record is a transaction record in which payment settlement has been completed before the current time, and the transaction settlement rule of each historical transaction record in the payment platform has been determined. In order to facilitate the distinction, the transaction settlement rule adopted by the historical transaction record to complete the transaction settlement is referred to as the actual transaction settlement rule, wherein the actual transaction settlement rule may also be considered as the transaction settlement rule actually configured by the payment platform for the historical transaction record.

S502. Select, from a plurality of historical transaction records, a historical transaction record sample to be trained.

The historical transaction record sample is a historical transaction record of the historical transaction record that is not used as a training sample to train the RNN model. For the sake of distinction, the historical transaction record to be trained selected from the historical transaction record that has never been used as the training sample is called A historical transaction record, wherein the historical transaction record is the currently selected target historical transaction record to be trained on the RNN model.

S503. Enter historical transaction information of the at least one attribute included in the historical transaction record sample into the RNN model to be trained, and obtain the historical transaction record sample output by the RNN model separately from the preset multiple different transactions. The probability of matching the settlement rules.

Among them, a plurality of different transaction settlement rules are preset in the RNN model to be trained.

S504, determining whether a matching probability between the historical transaction record sample output by the RNN model and the actual transaction settlement rule of the historical transaction record sample is greater than a preset threshold, and if yes, executing step S505; if not, adjusting the RNN The parameter settings in the model are returned to step S503.

The preset threshold may be set as needed. For example, the preset threshold may be 0.9.

It can be understood that if the parameter setting in the RNN model is reasonable, the matching probability between the historical transaction record sample outputted in the RNN model and the actual transaction settlement rule of the historical transaction record sample is larger. For example, greater than the preset threshold.

If the probability of matching between the historical transaction record sample output by the RNN model and the actual transaction settlement rule of the historical transaction record sample is low, it indicates that the parameter set in the RNN model is unreasonable, resulting in the output of the historical transaction record and The deviation of the matching probability between each transaction settlement rule is large. In this case, after adjusting the parameters in the RNN model, the historical transaction sample record is reused to train the RNN model until the matching probability between the historical transaction record sample output by the RNN model and its actual transaction settlement rule. Exceeded the preset threshold.

S505. Detect whether there is a historical transaction record in the plurality of historical transaction records that has not been trained as a training sample, and if yes, return to step S502; if not, the training ends, and the training obtains an RNN model as a rule matching model. .

After the plurality of historical transaction records are sequentially trained as training samples for the RNN model, the trained RNN model can be used as a model for calculating the matching probability of the matching between the transaction record and the different transaction settlement rules.

It should be noted that FIG. 5 is only for the convenience of understanding, and the RNN model is taken as an example to introduce the training rule matching model process. However, when the rule matching model is another network model or classifier, the training process is similar, and no longer Narration. At the same time, FIG. 5 is an example of a training process. However, in practical applications, the process of training the rule matching model may have other possible situations, which are not limited herein.

It can be understood that, considering that the payment platform often adds new transaction business scenarios due to new contracted merchants and contracted merchants changing corresponding transaction settlement rules, etc., in this case, if the historical transaction record is generated at a time distance The current time is longer, then the rule matching model trained based on historical transaction records cannot accurately reflect the degree of matching between different transaction records and transaction settlement rules. In order to make the trained rule matching model more accurately determine the transaction settlement rules matched by different transaction records, after the rule matching model is trained based on the historical transaction records, the rule matching model may be updated periodically.

For example, according to the set collection period, obtain a plurality of latest historical transaction records collected in the latest collection period before the current time, wherein the collection period may need to be set, for example, one day may be an acquisition period, or 4 Hours are an acquisition cycle, etc. For the sake of distinction, the historical transaction records collected in the most recent collection period before the current time are referred to as the latest historical transaction records, and the latest historical transaction records may belong to various transaction information of different attributes. In this way, when the retraining time corresponding to the rule matching model is reached, the rule matching model is trained to update the plurality of transaction information included in the latest historical transaction records collected in the most recent collection period. Rule matching model. The process of retraining the rule matching model by using the plurality of latest historical transaction records is similar to the process of training the rule matching model by using the historical transaction record, and details are not described herein.

The retraining time may be determined according to a retraining period corresponding to the rule matching model, so as to implement periodic updating of the rule matching model.

Optionally, when the model retraining time corresponding to the rule matching model is reached, the historical transaction record collected before the most recent collection period and the latest historical transaction record collected in the latest collection period may be used to match the rule matching model. Retrain to update the rule matching model. Wherein, when the rule matching model is retrained, the ratio of the historical transaction record and the latest transaction record before the latest collection period can be set as needed, and is not limited herein.

It can be understood that, in the case that the rule matching model needs to be updated, since the time required for the training rule to match the model is long, in order to avoid that the rule matching model has not completed the training, the rule matching model cannot be used to determine and be processed. In the embodiment of the present application, at least two pre-trained rule matching models may be preset, and the model retraining time of the at least two rule matching models is not completely the same, so that When the rule matching model needs to be called to determine the matching degree between the transaction record and the preset multiple transaction settlement rules, the rule matching model that is not currently in the training state may be selected from the preset at least two rule matching models.

In order to facilitate understanding, in the process of transaction settlement of transaction records, it is necessary to determine the transaction fee and other related expenses from the transaction amount of the transaction record according to the transaction settlement rule, and then finally complete the transaction settlement of the transaction record as an example. For example, refer to FIG. 6 , which is a schematic flowchart of a transaction settlement method applied to an application scenario, and the method in this embodiment may include:

S601: Obtain a transaction record to be settled.

The transaction record may include transaction information of at least two attributes including the payee of the transaction and the transaction amount of the transaction. In this embodiment, since the deduction is required from the transaction amount of the payee in the transaction settlement process, at least the specific values of the two attributes of the payee and the transaction amount need to be recorded in the transaction record.

For example, the payee of the transaction may be the merchant or user responsible for collecting the payment; the merchant or user accepting the transfer; or the user who withdraws the money, and the like. Correspondingly, the transaction amount can be the payment amount paid during the transaction, the transfer amount of the transfer, or the cash withdrawal amount, etc.

It can be understood that the transaction record can record other information related to the transaction, in addition to the transaction record recorded under the two attributes of the transaction and the transaction amount, for example, the aforementioned transaction. The transaction information of the payer, the payment channel of the transaction, the product type of the transaction, and the like.

S602. Determine, according to the transaction information of the at least two attributes included in the transaction record, and use a rule matching model obtained by the pre-training to determine a matching probability between the transaction record and the preset multiple transaction settlement rules.

S603. Determine, according to the matching probability between the transaction record and the preset plurality of different transaction settlement rules, determine, from the preset plurality of different transaction settlement rules, the target transaction settlement with the highest degree of matching with the transaction record. rule.

For the steps S602 and S603, reference may be made to related descriptions of the foregoing embodiments, and details are not described herein again.

S604. Calculate the deduction amount to be deducted from the transaction record according to the target transaction settlement rule.

In this embodiment, the transaction settlement rule can be understood as a charging rule that the payment platform performs charging for the transaction record. According to the charging rule, the deduction amount that needs to be deducted from the transaction amount for the transaction record can be determined.

For example, if the transaction settlement rule is the product of the deduction amount equal to 0.6% of the transaction amount and the billing rate, the amount obtained by multiplying the transaction amount by 0.6% is the deduction amount; for example, for billing If the amount is 1 yuan per transaction, the deduction amount for the transaction record is 1 yuan.

S605, the settlement amount obtained by subtracting the transaction amount of the transaction record from the deduction amount is deposited into the account designated by the payee of the transaction record.

For example, if the transaction amount is 100 yuan and the deduction amount is 1 yuan, the remaining 99 yuan is deposited into the account specified by the payee in the transaction record to realize settlement of the transaction based on the payment platform.

It can be understood that after calculating the deduction amount required for the transaction record, based on the transaction amount of the transaction record and the deduction amount, the transaction settlement of the transaction record can be completed.

S606, generating a transaction estimation voucher for the transaction record, and storing the transaction estimation voucher.

In the embodiment of the present application, the transaction estimation voucher is equivalent to the charging estimation voucher for estimating the transaction record. At least the transaction information including the payee and the transaction amount included in the transaction record, the transaction information, the target settlement rule according to the transaction settlement, and the transaction settlement corresponding to the transaction record are stored in the transaction estimation voucher. The amount is the deduction amount.

S607. When receiving the deviation compensation amount input by the user for the transaction estimation voucher, storing a correspondence between the deviation compensation amount and the payee of the transaction record, so as to receive the payment according to the deviation compensation amount. The party will compensate for the settlement amount of income after the current time.

Similar to the previous embodiment, after the transaction estimation voucher of the transaction record is stored in the server, the user can analyze the target transaction settlement rule corresponding to the transaction record and whether the deduction amount is correct according to the transaction estimation voucher. If the transaction settlement rule is incorrect, after the user analyzes the transaction settlement rule that actually matches the transaction record, the transaction settlement rule corresponding to the transaction record is reconfigured in the payment platform.

Meanwhile, if the deduction amount is incorrect, the actual deduction amount to be deducted from the transaction record may be recalculated, and the deviation compensation amount is determined according to the actual deduction amount and the deduction amount deducted in step S605. For example, the deviation compensation amount may be: using the target transaction settlement rule to calculate the difference between the deduction amount and the actual deduction amount. Correspondingly, after the deviation compensation amount is input to the server of the payment platform, the server deducts or increases the settlement amount of the income received by the payee after the current time according to the deviation compensation amount, to the payee Make amount compensation.

Similar to the previous embodiment, the step S606 and the step S607 are also optional steps, and the steps S606 and S607 may be selected or not executed according to actual needs.

In order to facilitate the understanding of the execution flow of the transaction settlement method of the present application, the transaction settlement process of the present application will be described below in conjunction with the execution flow of FIG. In the process, the payment platform is described as an example of a third-party payment platform, but the same applies to the payment platform.

In the transaction settlement process in FIG. 7, the transaction flow is started on the left side. As shown in FIG. 7, the third-party payment platform obtains the transaction flow, and the transaction flow is composed of a plurality of transaction records collected by the third-party payment platform at different times.

Because the third-party payment platform may lose data due to asynchronous delay accounting, rollback, etc., it is necessary to check the data in the transaction flow through the snapshot compensation module.

The data in the transaction stream is compensated by the snapshot compensation module, and then stored in the service snapshot module. The service snapshot module stores a plurality of transaction records to be processed involved in the transaction stream.

The parser analyzes the transaction information of different attributes in the transaction record, and determines whether the billing configuration module of the third-party platform stores the transaction settlement rule having a mapping relationship with the transaction record.

On one hand, if the parser parses out the transaction settlement rule configured by the third-party payment platform with the transaction record, the transaction settlement rule of the transaction record is obtained from the billing configuration module, and the transaction settlement rule is calculated based on the transaction settlement rule of the transaction record. The billing of the transaction record (ie, the deduction amount) and the settlement of the transaction for the transaction record is completed. After the transaction settlement is completed, a billing document is generated. As shown in FIG. 7, the "accounting" connected to the billing configuration module indicates the transaction record of the billing according to the transaction settlement rules configured in the billing configuration module.

The billing voucher may include a billing and a transaction settlement rule corresponding to the transaction record. The transaction record included in the billing voucher and the transaction settlement rule of the transaction record can be used as the transaction transaction rule for training the rule matching model and the transaction settlement rule of the recent transaction record. As shown in FIG. 7, the billing certificate can be As a sample input of the sampler corresponding to the learner, the learner trained in the historical transaction record is trained to achieve the purpose of updating the learner. In Figure 7, the learner refers to a rule matching model that requires repeated training.

In FIG. 7, the third-party payment platform is exemplified by setting two learners, which are trained by using data corresponding to historical transaction records in advance.

On the other hand, if the parser resolves the transaction settlement rule in which the third-party payment platform is not configured with the transaction record, a learner that is not currently in the training state is selected from the at least two learners obtained in advance. The selected learner determines the degree of matching between the transaction record and the preset multiple transaction settlement rules, thereby determining the target transaction settlement rule with the highest degree of matching with the transaction record. The transaction record is then billed and the transaction is settled using the target transaction settlement rules. The "billing" (the third "billing" from left to right) connected to the billing configuration module by a dotted line in FIG. 7 by a broken line indicates a transaction record not configured with a transaction settlement rule.

At the same time, for the transaction record that learns the target transaction settlement rule through the learner, after completing the transaction settlement of the transaction record, a billing estimation voucher is also generated for the transaction record, as shown in FIG. The estimation module stores the accounting estimation voucher corresponding to all the transaction records in which the transaction settlement rule is completed by the target transaction settlement rule learned by the learner.

By artificially supervising the billing estimated voucher, the transaction record with the billing error can be manually analyzed, and the amount of the offset compensation required for the transaction record with the billing error can be manually determined, so as to subsequently receive the transaction record. The amount is deducted or compensated in the payment. At the same time, according to the transaction record included in the billing estimation voucher, the user can analyze the transaction settlement rules actually required by the transaction record, and configure the transaction settlement rules actually needed for the transaction record in the billing configuration model.

A transaction settlement device provided by an embodiment of the present invention is described below. A transaction settlement device described below may be associated with a transaction settlement method described above.

For example, referring to FIG. 8, a schematic structural diagram of an embodiment of a transaction settlement device of the present application is shown. The device of this embodiment may include:

a record obtaining unit 801, configured to acquire a transaction record to be settled;

The rule matching unit 802 is configured to determine, by using a pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, where the rule matching model is based on multiple previous times The historical transaction record and the transaction settlement rules corresponding to the historical transaction record are trained;

The rule determining unit 803 is configured to determine, according to the matching degree of the transaction record and the preset plurality of transaction settlement rules, from the preset plurality of transaction settlement rules, determine the transaction settlement required for the transaction record Target transaction settlement rules;

The transaction settlement unit 804 is configured to perform transaction settlement on the transaction record based on the target transaction settlement rule.

The transaction record acquired by the record obtaining unit includes at least one attribute of transaction information;

The rule matching unit is configured to determine, according to transaction information of at least one attribute in the transaction record and a matching rule that is mined in a pre-trained rule matching model, and use the rule matching model to determine the transaction record. The degree of matching with the preset plurality of transaction settlement rules, wherein the matching rule includes: different transaction information included in each attribute, and the degree of influence of the transaction record on the preset multiple transaction settlement rules.

In an implementation manner, the apparatus further includes: a model training unit, configured to obtain the rule matching model by:

Obtaining a plurality of historical transaction records required for training and an actual transaction settlement rule of each historical transaction record, wherein the actual transaction settlement rule is a transaction settlement rule adopted by the historical transaction record to complete transaction settlement, and the historical transaction record includes Historical transaction information of at least one attribute;

Selecting, from the plurality of historical transaction records, a historical transaction record of the target to be trained;

Inputting historical transaction information of at least one attribute of the target historical transaction record into a rule matching model to be trained, and acquiring the target historical transaction records output by the rule matching model and different presets respectively The matching probability of the settlement rule;

And when the target matching probability output by the rule matching model is less than or equal to a preset threshold, adjusting a parameter setting in the rule matching model, and returning to perform the historical transaction of the at least one attribute of the target historical transaction record The information is input to an operation in the rule matching model to be trained, wherein the target matching probability is a matching probability between the target historical transaction record and the actual transaction settlement rule of the target historical transaction record;

When the target matching probability output by the rule matching model is not less than the preset threshold, returning to performing the operation of selecting the current historical transaction record to be trained from the plurality of historical transaction records until the There are no historical transaction records that have not been trained in the historical transaction records.

In one implementation, the apparatus further includes:

a data collecting unit, configured to acquire, according to a preset collection period, an latest historical transaction record in which the payment platform completes transaction settlement in the collection period and a transaction settlement rule corresponding to the latest historical transaction record;

a model retraining unit, configured to use the latest historical transaction record collected in the latest collection period before the current time and the transaction settlement rule corresponding to the latest historical transaction record when the model retraining time corresponding to the rule matching model is reached And training the rule matching model to update the rule matching model.

In an implementation manner, the rule matching unit includes:

a model selection sub-unit, configured to determine, from a plurality of pre-trained rule matching models, a rule matching model that is not currently in a training state;

The rule matching sub-unit is configured to determine, by using a rule matching model that is not currently in the training state, a degree of matching between the transaction record and the preset multiple transaction settlement rules.

Optionally, the rule determining unit is specifically configured to determine, from the preset plurality of transaction settlement rules, a target transaction settlement rule with the highest degree of matching with the transaction record.

Optionally, the transaction record acquired by the record obtaining unit includes at least two types of transaction information belonging to different attributes, and the at least two pieces of transaction information include: a payee of the transaction record and the transaction record Transaction amount;

The transaction settlement unit includes:

a billing subunit, configured to determine a deduction amount to be deducted from the transaction record based on the target transaction settlement rule;

And a settlement subunit, configured to reduce the transaction amount in the transaction record by the deduction amount, and deposit the settlement amount into the account specified by the payee of the transaction record.

In an implementation manner, the transaction settlement device further includes:

An estimated voucher generating unit, configured to generate a transaction estimation voucher for the transaction record after the transaction settlement unit performs transaction calculation on the transaction record, and store the transaction estimation voucher, the transaction estimation voucher The method includes: the at least two transaction information in the transaction record, a target transaction settlement rule corresponding to the transaction record, and a settlement amount of the transaction record for settlement of the transaction;

The amount compensation unit is configured to store, when receiving the deviation compensation amount input by the user for the transaction estimation voucher, a correspondence between the deviation compensation amount and the payee of the transaction record, so as to compensate the amount according to the deviation And compensate the settlement party for the settlement amount of the latest income after the current time.

On the other hand, the embodiment of the present application further provides a storage medium for storing program code, and the program code is used to execute the transaction settlement method described in any one of the above embodiments. In another aspect, embodiments of the present invention also provide a computer program product comprising instructions that, when executed on a computer, cause the computer to perform transaction settlement as described in any of the above embodiments.

It should be noted that each embodiment in the specification is described in a progressive manner, and each embodiment focuses on differences from other embodiments, and the same similar parts between the embodiments are referred to each other. can. For the device type embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment.

Finally, it should also be noted that in this context, relational terms such as first and second are used merely to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these entities. There is any such actual relationship or order between operations. Furthermore, the term "comprises" or "comprises" or "comprises" or any other variations thereof is intended to encompass a non-exclusive inclusion, such that a process, method, article, or device that comprises a plurality of elements includes not only those elements but also Other elements, or elements that are inherent to such a process, method, item, or device. An element defined by the phrase "comprising a ..." without further limitation does not exclude the presence of additional identical elements in the process, method, article, or device that comprises the element.

The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments are obvious to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but the scope of the invention is to be accorded

The above is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. These improvements and retouchings should also be considered. It is the scope of protection of the present invention.

Claims

A transaction settlement method, including:

Obtain transaction records to be settled;

Determining, by the pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, wherein the rule matching model is based on a plurality of historical transaction records before the current time and the history The trading settlement rules corresponding to the transaction records are trained;

Determining a target transaction settlement rule required for transaction settlement of the transaction record from the preset plurality of transaction settlement rules according to a matching degree of the transaction record with a preset plurality of transaction settlement rules;

Transaction settlement is performed on the transaction record based on the target transaction settlement rule.
The transaction settlement method according to claim 1, wherein the transaction record includes transaction information of at least one attribute;

Determining, by using a pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, including:

Determining, according to the transaction information of the at least one attribute in the transaction record and the matching rule excerpted in the pre-trained rule matching model, the degree of matching between the transaction record and the preset multiple transaction settlement rules, wherein The matching rule includes: different transaction information included in each attribute, and a degree of influence on matching the transaction record with the preset multiple transaction settlement rules.
The transaction settlement method according to claim 1, further comprising:

Obtaining a plurality of historical transaction records required for training and an actual transaction settlement rule of each historical transaction record, wherein the actual transaction settlement rule is a transaction settlement rule adopted by the historical transaction record to complete transaction settlement, and the historical transaction record includes Historical transaction information of at least one attribute;

Selecting, from the plurality of historical transaction records, a historical transaction record of the target to be trained;

Inputting historical transaction information of at least one attribute of the target historical transaction record into a rule matching model to be trained, and acquiring the target historical transaction record output by the rule matching model to be trained respectively and preset Matching probability of different transaction settlement rules;

And when the target matching probability output by the rule matching model to be trained is less than or equal to a preset threshold, adjusting a parameter setting in the rule matching model to be trained, and returning to perform performing at least the target historical transaction record An operation of inputting historical transaction information of an attribute into a rule matching model to be trained, wherein the target matching probability is a matching probability between the target historical transaction record and an actual transaction settlement rule of the target historical transaction record;

When the target matching probability output by the rule matching model to be trained is not less than the preset threshold, returning to performing the operation of selecting the current historical transaction record to be trained from the plurality of historical transaction records Until there are no historical transaction records that have not been trained in the plurality of historical transaction records.
The transaction settlement method according to claim 1 or 3, further comprising:

Obtaining, according to a preset collection period, an latest historical transaction record that completes transaction settlement in the collection period and a transaction settlement rule corresponding to the latest historical transaction record;

When the model retraining time corresponding to the rule matching model is reached, the latest historical transaction record collected in the latest collection period before the current time and the transaction settlement rule corresponding to the latest historical transaction record are used to match the rule matching model. Training is performed to update the rule matching model.
The transaction settlement method according to claim 4, wherein the pre-trained rule matching model is used to determine a degree of matching between the transaction record and a preset plurality of transaction settlement rules, including:

Determining a rule matching model that is not currently in a training state from a plurality of pre-trained rule matching models;

The degree of matching between the transaction record and the preset plurality of transaction settlement rules is determined by using a rule matching model that is not currently in the training state.
The transaction settlement method according to claim 1, wherein the transaction record includes at least two transaction information belonging to different attributes, and the at least two transaction information includes: a payee of the transaction record and the transaction record Amount of the transaction;

The transaction settlement of the transaction record based on the target transaction settlement rule comprises:

Determining a deduction amount to be deducted from the transaction record based on the target transaction settlement rule;

The settlement amount obtained by subtracting the transaction amount in the transaction record from the deduction amount is deposited in the account designated by the payee of the transaction record.
The transaction settlement method according to claim 6, after the transaction settlement of the transaction record based on the target transaction settlement rule, further comprising:

Generating a transaction estimation voucher for the transaction record, and storing the transaction estimation voucher, the transaction estimation voucher including: the at least two transaction information in the transaction record, and a target transaction corresponding to the transaction record The settlement rule and the settlement amount of the transaction record for settlement of the transaction;

And when a deviation compensation amount input by the user for the transaction estimation voucher is received, storing a correspondence between the deviation compensation amount and a payee of the transaction record, so as to receive the payment according to the deviation compensation amount The party will compensate for the settlement amount after the current time.
A transaction settlement device comprising:

a record acquisition unit for obtaining a transaction record to be settled;

a rule matching unit, configured to determine, by using a pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, wherein the rule matching model is based on multiple history before the current time Transaction records and training settlement rules corresponding to the historical transaction records;

a rule determining unit, configured to determine, according to a matching degree of the transaction record and a preset plurality of transaction settlement rules, from the preset plurality of transaction settlement rules, determining a transaction settlement required for the transaction record Target transaction settlement rules;

a transaction settlement unit, configured to perform transaction settlement on the transaction record based on the target transaction settlement rule.
The transaction settlement device according to claim 8, wherein the transaction record acquired by the record acquisition unit includes transaction information of at least one attribute;

The rule matching unit is configured to determine, according to the transaction information of the at least one attribute in the transaction record and the matching rule excerpted in the pre-trained rule matching model, to determine the transaction record and the preset multiple The matching degree of the transaction settlement rule, wherein the matching rule includes: different transaction information included in each attribute, and the degree of influence of the transaction record on the preset multiple transaction settlement rules.
The transaction settlement device of claim 8 further comprising:

a data collecting unit, configured to acquire, according to a preset collection period, an latest historical transaction record that completes transaction settlement in the collection period and a transaction settlement rule corresponding to the latest historical transaction record;

a model retraining unit, configured to use the latest historical transaction record collected in the latest collection period before the current time and the transaction settlement rule corresponding to the latest historical transaction record when the model retraining time corresponding to the rule matching model is reached And training the rule matching model to update the rule matching model.
The transaction settlement device according to claim 10, wherein the rule matching unit comprises:

a model selection sub-unit, configured to determine, from a plurality of pre-trained rule matching models, a rule matching model that is not currently in a training state;

The rule matching sub-unit is configured to determine, by using a rule matching model that is not currently in the training state, a degree of matching between the transaction record and the preset multiple transaction settlement rules.
The transaction settlement device according to claim 8, wherein the transaction record acquired by the record acquisition unit includes at least two transaction information belonging to different attributes, and the at least two transaction information includes: receipt of the transaction record The amount of the transaction and the transaction amount of the transaction record;

The transaction settlement unit includes:

a billing subunit, configured to determine a deduction amount to be deducted from the transaction record based on the target transaction settlement rule;

And a settlement subunit, configured to reduce the transaction amount in the transaction record by the deduction amount, and deposit the settlement amount into the account specified by the payee of the transaction record.
The transaction settlement device according to claim 12, wherein the transaction settlement device further comprises:

An estimated voucher generating unit, configured to generate a transaction estimation voucher for the transaction record after the transaction settlement unit performs transaction settlement on the transaction record, and store the transaction estimation voucher, the transaction estimation voucher The method includes: the at least two transaction information in the transaction record, a target transaction settlement rule corresponding to the transaction record, and a settlement amount of the transaction record for settlement of the transaction;

The amount compensation unit is configured to store, when receiving the deviation compensation amount input by the user for the transaction estimation voucher, a correspondence between the deviation compensation amount and the payee of the transaction record, so as to compensate the amount according to the deviation , the settlement amount of the payee after the current time is compensated.
A server that includes:

Processor and memory;

Wherein the processor is configured to execute a program stored in the memory;

The memory is for storing a program, the program being at least for:

Obtain transaction records to be settled;

Determining, by the pre-trained rule matching model, a degree of matching between the transaction record and a preset plurality of transaction settlement rules, wherein the rule matching model is based on a plurality of historical transaction records before the current time and the history The trading settlement rules corresponding to the transaction records are trained;

Determining a target transaction settlement rule required for transaction settlement of the transaction record from the preset plurality of transaction settlement rules according to a matching degree of the transaction record with a preset plurality of transaction settlement rules;

Transaction settlement is performed on the transaction record based on the target transaction settlement rule.
A transaction settlement method, the method being applied to a server, the method comprising:

The server performs the transaction settlement method according to any one of claims 1 to 7.
A storage medium for storing program code for performing the transaction settlement method of any one of claims 1-7. .
A computer program product comprising instructions which, when run on a computer, cause the computer to perform the transaction settlement method of any of claims 1-7.