Transaction Matching Method, Device, Equipment, Medium and Program Product

Abstract

The present disclosure provides a transaction matching method, apparatus, device, medium, and program product, relating to the field of big data technology. Among them, the method includes: obtaining fund supply information, where the fund supply information includes m fund providers and the investment amount of each fund provider; obtaining borrowing demand information, where the borrowing demand information includes n borrowers and the borrowing amount and borrowing interest rate of each borrower; based on the sparsity constraint condition and the consistency constraint condition, matching the fund supply information and the borrowing demand information through the method of operations research optimization to obtain a matching result matrix, where the element in the i-th row and j-th column of the matching result matrix represents the transaction amount between the i-th fund provider and the j-th borrower. Based on the transaction amounts in the matching result matrix, fund transactions are carried out among the m fund providers and the n borrowers.

Description

Technical Field

[0001] The present disclosure relates to the field of big data technology, and can be applied to the field of fintech. Specifically, it relates to a transaction matching method, device, equipment, medium and program product. Background Art

[0002] In financing business scenarios such as the interbank market, there are often transaction matching problems between multiple fund providers and multiple borrowers. For example, Fund A provides a capital contribution of x, Fund B provides a capital contribution of y, and Fund C provides a capital contribution of z. At the same time, different borrowers have different borrowing requirements. For example, the borrowing amount of Institution A is a, and the borrowing amount of Institution B is b, and the borrowing interest rates promised by different institutions are different. In such a scenario, it is necessary to design a transaction matching plan to match the needs of lenders and borrowers. Moreover, when designing a transaction matching plan, there are usually some business requirements, such as consistency constraints and sparsity constraints. Among them, the consistency constraint means that in the final transaction matching result, the income returns of each fund provider should be as small as possible to avoid affecting the enthusiasm of fund providers to participate in financing, and the sparsity constraint means that the borrowing of each borrower is concentrated in a small number of fund providers as much as possible, which can simplify the formalities in the transaction process, etc.

[0003] In the prior art, the design of the transaction matching plan is mainly completed manually. Among them, in order to meet various business requirements, usually manual trial calculations are carried out and continuously adjusted and optimized. This not only has low efficiency and consumes human resources, but also the final matching plan is usually not the optimal plan. Moreover, with the increase in the scale of the financing business, the number of fund providers and borrowers increases, and it is increasingly difficult to manually design a transaction matching plan to meet the timeliness and high-quality requirements of transaction matching. Summary of the Invention

[0004] In view of this, the present disclosure provides a transaction matching method, device, equipment, medium and program product that can automatically provide a transaction matching plan according to operations research optimization.

[0005] In the first aspect of the embodiments of the present disclosure, a transaction matching method is provided. The method includes: obtaining fund supply information, where the fund supply information includes m fund providers and the capital contribution of each fund provider, and m is an integer greater than 1; obtaining borrowing demand information, where the borrowing demand information includes n borrowers and the borrowing amount and borrowing interest rate of each borrower, and n is an integer greater than 1; based on the sparsity constraint condition and the consistency constraint condition, matching the fund supply information and the borrowing demand information by means of operations research optimization to obtain a matching result matrix, where the matching result matrix is an m*n matrix, and the element in the i-th row and j-th column of the matrix represents the transaction amount between the i-th fund provider and the j-th borrower; and based on the transaction amounts in the matching result matrix, fund transactions are conducted among the m fund providers and the n borrowers. Among them, the sparsity constraint condition includes that each borrower can obtain funds from at most S fund providers, where S is an integer greater than 1 and less than m. The consistency constraint condition includes that the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to a predetermined interest rate difference threshold, where the m weighted interest rates correspond one-to-one with the m fund providers, and the weighted interest rate corresponding to each fund provider is the weighted average of the borrowing interest rates of all borrowers with which the fund provider matches the transaction amount, and the weight of the weighted average is the ratio of the transaction amount to the investment amount of the fund provider.

[0006] According to an embodiment of the present disclosure, the matching of the fund supply information and the borrowing demand information based on the sparsity constraint condition and the consistency constraint condition through operational optimization to obtain the matching result matrix includes: matching the fund supply information and the borrowing demand information based on the sparsity constraint condition to obtain a first matching matrix; calculating the m weighted interest rates based on the first matching matrix; and when the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to the interest rate difference threshold, determining the first matching matrix as the matching result matrix.

[0007] According to an embodiment of the present disclosure, the method further includes, when the difference between the maximum value and the minimum value among the m weighted interest rates is greater than the interest rate difference threshold, cyclically performing the following operations until the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold: determining a first fund provider corresponding to the maximum value and a second fund provider corresponding to the minimum value among the m weighted interest rates; obtaining local fund supply information and local fund demand information; wherein, the local fund supply information includes the investment amounts of the first fund provider and the second fund provider; the local fund demand information includes all borrowers to be optimized and the local borrowing amounts of each borrower to be optimized, wherein the borrower to be optimized is a borrower who matches a transaction amount with at least one of the first fund provider and the second fund provider in the first matching matrix, and the local borrowing amount is the sum of the transaction amounts matched by the borrower to be optimized with at least one of the first fund provider and the second fund provider in the first matching matrix; matching the local fund supply information and the local fund demand information based on the local consistency condition to obtain a second matching matrix, wherein the number of rows of the second matching matrix is 2, and each element represents the transaction amount between one of the first fund provider and the second fund provider and a borrower to be optimized; using the transaction amounts in the second matching matrix to correspondingly replace the transaction amounts of the first fund provider and the second fund provider in the first matching matrix to update the first matching matrix; recalculating the m weighted interest rates based on the updated first matching matrix to update the m weighted interest rates; comparing the difference between the maximum value and the minimum value among the updated m weighted interest rates with the interest rate difference threshold. Wherein, the local consistency condition includes that the difference between the weighted interest rate corresponding to the first fund provider and the weighted interest rate corresponding to the second fund provider is less than or equal to the interest rate difference threshold.

[0008] According to an embodiment of the present disclosure, the obtaining of the local fund supply information and the local fund demand information includes: extracting the rows corresponding to the first fund provider and the second fund provider from the first matching matrix to form a first low-dimensional matrix; removing the columns with all column elements being zero from the first low-dimensional matrix to obtain a second low-dimensional matrix; and extracting the local fund demand information and the local fund demand information from the second low-dimensional matrix.

[0009] According to an embodiment of the present disclosure, matching the local fund supply information and the local fund demand information based on the local consistency condition to obtain the second matching matrix includes: using the consistency solver to match the local fund supply information and the local fund demand information, and solving to obtain the second matching matrix; wherein, the objective function of the consistency solver includes minimizing the weighted value of the consistency penalty function and the sparse penalty function, and the consistency penalty function is a function obtained based on the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider.

[0010] According to an embodiment of the present disclosure, the weight of the consistency penalty function in the objective function of the consistency solver is 1, and the weight of the sparse penalty function is iteratively determined in the following manner after setting an initial value: obtaining the solution matrix obtained by the consistency solver matching the local fund supply information and the local fund demand information; comparing the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider obtained based on the solution matrix with the interest rate difference threshold to obtain a comparison result; when the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is greater than the interest rate difference threshold, reducing the weight of the sparse penalty function in the objective function of the consistency solver to update the objective function of the consistency solver, and using the updated consistency solver to rematch the local fund supply information and the local fund demand information.

[0011] According to an embodiment of the present disclosure, reducing the weight of the sparse penalty function in the objective function of the consistency solver includes: reducing the weight of the sparse penalty function in the objective function of the consistency solver in an exponentially decaying form.

[0012] According to an embodiment of the present disclosure, matching the fund supply information and the borrowing demand information based on the sparsity constraint condition to obtain the first matching matrix includes: using the sparsity solver to match the fund supply information and the borrowing demand information, and solving to obtain the first matching matrix; wherein, the objective function of the sparsity solver includes minimizing the sparse penalty function.

[0013] In a second aspect of the embodiments of the present disclosure, a transaction matching device is provided. The device includes: a first acquisition module, a second acquisition module, a matching module, and a transaction module.

[0014] The first acquisition module is used to acquire fund supply information, where the fund supply information includes m fund providers and the contribution amount of each fund provider, and m is an integer greater than 1.

[0015] The second acquisition module is used to acquire borrowing demand information, where the borrowing demand information includes n borrowers and the borrowing amount and borrowing interest rate of each borrower, and n is an integer greater than 1.

[0016] The matching module is used to match the fund supply information and the borrowing demand information by means of operations research optimization based on the sparsity constraint condition and the consistency constraint condition, and obtain a matching result matrix, where the matching result matrix is an m*n matrix, and the element in the i-th row and j-th column represents the transaction amount between the i-th fund provider and the j-th borrower; where the sparsity constraint condition includes that each borrower can obtain funds from at most S fund providers, and S is an integer greater than 1 and less than m; the consistency constraint condition includes that the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to a predetermined interest rate difference threshold, where the m weighted interest rates correspond one-to-one with the m fund providers, and the weighted interest rate corresponding to each fund provider is the weighted average of the borrowing interest rates of all borrowers that match the transaction amount with the fund provider, and the weight of the weighted average is the ratio of the transaction amount to the investment amount of the fund provider.

[0017] The transaction module is used to conduct fund transactions between the m fund providers and the n borrowers based on the transaction amounts in the matching result matrix.

[0018] According to an embodiment of the present disclosure, the matching module includes: a first matching sub-module and a tuning sub-module.

[0019] The first matching sub-module is used to: match the fund supply information and the borrowing demand information based on the sparsity constraint condition, and obtain a first matching matrix;

[0020] The tuning sub-module is used to: calculate the m weighted interest rates based on the first matching matrix; and when the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to the interest rate difference threshold, determine the first matching matrix as the matching result matrix.

[0021] According to an embodiment of the present disclosure, the tuning sub-module is further configured to: when the difference between the maximum value and the minimum value among the m weighted interest rates is greater than the interest rate difference threshold, loop to perform the following operations until the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold: determine the first fund provider corresponding to the maximum value and the second fund provider corresponding to the minimum value among the m weighted interest rates; obtain local fund supply information and local fund demand information; wherein, the local fund supply information includes the contribution amounts of the first fund provider and the second fund provider; the local fund demand information includes all borrowers to be optimized and the local borrowing amounts of each borrower to be optimized, wherein the borrower to be optimized is a borrower in the first matching matrix that matches a transaction amount with at least one of the first fund provider and the second fund provider, and the local borrowing amount is the sum of the transaction amounts that the borrower to be optimized matches with at least one of the first fund provider and the second fund provider in the first matching matrix; match the local fund supply information and the local fund demand information based on the local consistency condition to obtain a second matching matrix, wherein the number of rows of the second matching matrix is 2, and each element represents the transaction amount between one of the first fund provider and the second fund provider and a borrower to be optimized; use the transaction amounts in the second matching matrix to correspondingly replace the transaction amounts of the first fund provider and the second fund provider in the first matching matrix to update the first matching matrix; recalculate the m weighted interest rates based on the updated first matching matrix to update the m weighted interest rates; compare the difference between the maximum value and the minimum value among the updated m weighted interest rates with the interest rate difference threshold. Wherein, the local consistency condition includes that the difference between the weighted interest rate corresponding to the first fund provider and the weighted interest rate corresponding to the second fund provider is less than or equal to the interest rate difference threshold.

[0022] According to an embodiment of the present disclosure, the obtaining of the local fund supply information and the local fund demand information includes: extracting the rows corresponding to the first fund provider and the second fund provider from the first matching matrix to form a first low-dimensional matrix; removing the columns with all column elements being zero from the first low-dimensional matrix to obtain a second low-dimensional matrix; and extracting the local fund demand information and the local fund demand information from the second low-dimensional matrix.

[0023] According to an embodiment of the present disclosure, the tuning sub-module includes a consistency solver. Wherein, the tuning sub-module is further configured to: use the consistency solver to match the local fund supply information and the local fund demand information, and solve to obtain the second matching matrix. Wherein, the objective function of the consistency solver includes minimizing the weighted value of a consistency penalty function and a sparsity penalty function. Wherein, the consistency penalty function is a function obtained based on the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider.

[0024] According to an embodiment of the present disclosure, the tuning sub-module further includes a consistency solver construction sub-module. The consistency solver construction module is configured to construct the consistency solver, including setting the weight of the consistency penalty function in the objective function of the consistency solver to 1, and setting the weight of the sparsity penalty function to be iterated in the following manner after initialization: obtaining the solution matrix obtained by the consistency solver matching the local fund supply information and the local fund demand information; comparing the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider obtained based on the solution matrix with the interest rate difference threshold to obtain a comparison result; when the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is greater than the interest rate difference threshold, reducing the weight of the sparsity penalty function in the objective function of the consistency solver to update the objective function of the consistency solver, and using the updated consistency solver to re-match the local fund supply information and the local fund demand information.

[0025] According to an embodiment of the present disclosure, the consistency solver construction sub-module is specifically configured to: when the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is greater than the interest rate difference threshold, reduce the weight of the sparsity penalty function in the objective function of the consistency solver in an exponentially decaying form.

[0026] According to an embodiment of the present disclosure, the first matching sub-module further includes a sparsity solver. Wherein, the first matching sub-module is further configured to: use the sparsity solver to match the fund supply information and the borrowing demand information, and solve to obtain the first matching matrix; wherein, the objective function of the sparsity solver includes minimizing the sparsity penalty function.

[0027] In a third aspect of the embodiments of the present disclosure, an electronic device is provided. The electronic device includes: one or more processors and a memory. The memory is used to store one or more computer programs. Wherein, the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0028] In a fourth aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium, on which a computer program or instructions are stored, and when the computer program or instructions are executed by a processor, the steps of the above method are implemented.

[0029] In a fifth aspect of the embodiments of the present disclosure, there is provided a computer program product, including a computer program or instructions, and when the computer program or instructions are executed by a processor, the steps of the above method are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Through the following description of the embodiments of the present disclosure with reference to the drawings, the above and other objects, features, and advantages of the present disclosure will become clearer. In the drawings:

[0031] Figure 1 Schematically shows an application scenario of a transaction matching method, device, equipment, medium, and program product according to an embodiment of the present disclosure;

[0032] Figure 2 Schematically shows a flowchart of a transaction matching method according to an embodiment of the present disclosure;

[0033] Figure 3 Schematically shows a flowchart of a transaction matching method according to another embodiment of the present disclosure;

[0034] Figure 4 Schematically shows a flowchart of optimizing a first matrix in a transaction matching method according to another embodiment of the present disclosure;

[0035] Figure 5 Schematically shows a flowchart of a transaction matching method according to still another embodiment of the present disclosure;

[0036] Figure 6 Schematically shows a flowchart of constructing a consistency solver in a transaction matching method according to still another embodiment of the present disclosure;

[0037] Figure 7 Schematically shows a block diagram of a transaction matching device according to an embodiment of the present disclosure;

[0038] Figure 8 Schematically shows a block diagram of a matching module according to an embodiment of the present disclosure; and

[0039] Figure 9 Schematically shows a block diagram of an electronic device suitable for implementing the transaction matching method according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the following detailed description, for the sake of explanation, numerous specific details are set forth in order to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments can also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts of the present disclosure.

[0041] The terms used herein are merely for describing specific embodiments and are not intended to limit the present disclosure. The terms "including", "comprising", etc. used herein indicate the presence of the described features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0042] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0043] In the case of using expressions such as "at least one of A, B, and C, etc.", generally, it should be interpreted according to the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include, but is not limited to, a system having only A, only B, only C, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.).

[0044] To overcome the problems of low efficiency in manually designing transaction matching schemes and difficulty in finding optimal schemes in the prior art, embodiments of the present disclosure provide a transaction matching method, apparatus, device, medium, and program product. Through the method of operations research optimization, on the premise of ensuring the satisfaction of the sparsity constraint and consistency constraint of business requirements, a transaction matching scheme (also referred to as a transaction component) can be automatically and efficiently output.

[0045] Specifically, the transaction matching scheme provided through the method of operations research optimization can be modeled as a matrix , where the element represents the transaction amount between the i-th fund provider and the j-th borrower.

[0046] According to the fund matching requirements in the transaction, it can be easily obtained that the sum of the elements in each row of the matrix should be equal to or infinitely close to the investment amount of the corresponding fund provider in that row, and the sum of the elements in each column should be equal to or infinitely close to the borrowing amount of the corresponding borrower in that column.

[0047] The sparsity constraint in the business requirements is to ensure the transaction execution efficiency, and the orders of each borrower cannot be split too thinly. In this regard, corresponding sparsity constraint conditions can be constructed in operations research optimization. For example, it is required to solve the matrix such that the number of non-zero elements in each column is less than or equal to S, where S can be specified manually (e.g., specified by the trading department, usually taken as 3 or 4).

[0048] The consistency constraint in the business requirements is to ensure that the return on investment of each investor is as consistent as possible to ensure fair distribution. In this regard, corresponding consistency constraint conditions can be constructed in operations research optimization. For example, it is required that the difference between the weighted interest rates of different investors obtained from the solution matrix (the definition and calculation process are referred to below) be limited to a predetermined interest rate difference threshold (e.g., 2 basis points).

[0049] From the above discussion, it can be seen that when manually designing a transaction matching scheme, if there are many investors and borrowers, and the capital scale is large, and at the same time at least the above two business constraint requirements need to be met, it will inevitably result in the consumption of a large amount of human resources, and it is often difficult to obtain satisfactory results.

[0050] In view of this, the embodiments of the present disclosure can perform transaction matching based on the above-mentioned sparsity constraint conditions and consistency constraint conditions through operations research optimization. Among them, necessary data is automatically collected from the business system, and relevant data is used to solve a transaction matching scheme that simultaneously satisfies the above two constraint conditions, so as to save human resources and improve the business level.

[0051] In addition, in the actual business operation, the transaction frequency is high. Sometimes it is necessary to propose a qualified transaction matching scheme within a limited time. When performing transaction matching, not only must it be able to solve, but it must also have a relatively fast solution speed to be practical. Therefore, in some embodiments, when performing transaction matching based on the above-mentioned sparsity constraint conditions and consistency constraint conditions through operations research optimization, the efficiency problem of model solution will also be considered to design an optimized matching scheme. For example, without considering the consistency constraint conditions, first solve according to the sparsity constraint conditions, and then check whether the solution result meets the consistency constraint conditions. If it meets, the solution result can be directly output. If it does not meet, then locally optimize the solution result according to the consistency constraint conditions, which can significantly improve the solution efficiency of the transaction matching scheme.

[0052] Figure 1 Schematically shows the application scenarios of the transaction matching method, device, equipment, medium and program product of an embodiment of the present disclosure.

[0053] Such as Figure 1As shown, the application scenario 100 according to this embodiment may include at least one terminal device (three are shown in the figure, terminal devices 101, 102, and 103), a network 104, and a server 105. The network 104 is a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0054] Users can use the terminal devices 101, 102, 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications may be installed on the terminal devices 101, 102, 103, such as shopping applications, government affairs applications, financial applications, instant messaging tools, email clients, social platform software, etc. (only for examples). The terminal devices 101, 102, 103 may be various electronic devices with a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, and desktop computers, etc.

[0055] The server 105 may be a server providing various services. The server 105 may execute the transaction matching method of the embodiments of the present disclosure.

[0056] For example, users can send instructions for transaction matching to the server 105 through the terminal devices 101, 102, 103. The server 105 may perform analysis and processing of transaction matching according to the user instructions, and feedback the transaction matching result or the transaction execution result to the terminal device.

[0057] It should be noted that the transaction matching method provided by the embodiments of the present disclosure may generally be executed by the server 105. Correspondingly, the transaction matching device, equipment, medium, and program product provided by the embodiments of the present disclosure may generally be set in the server 105. The transaction matching method provided by the embodiments of the present disclosure may also be executed by a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and / or the server 105. Correspondingly, the transaction matching device, equipment, medium, and program product provided by the embodiments of the present disclosure may also be set in a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and / or the server 105.

[0058] It should be understood that Figure 1 the numbers of terminal devices, networks, and servers in

[0059] Figure 2 are merely illustrative. According to the implementation requirements, there may be any number of terminal devices, networks, and servers.

[0060] As Figure 2 shown, the transaction matching method of this embodiment may include operations S201 to S204.

[0061] In operation S201, obtain the fund supply information, where the fund supply information includes m fund providers and the amount contributed by each fund provider, and m is an integer greater than 1.

[0062] In operation S202, obtain the borrowing demand information, where the borrowing demand information includes n borrowers and the borrowing amount and borrowing interest rate of each borrower, and n is an integer greater than 1.

[0063] In operation S203, based on the sparsity constraint condition and the consistency constraint condition, match the fund supply information and the borrowing demand information through the operation research optimization method to obtain a matching result matrix, where the matching result matrix is an m*n matrix, and the element in the i-th row and j-th column represents the transaction amount between the i-th fund provider and the j-th borrower. This matching result matrix is the result of the required transaction matching scheme.

[0064] The sparsity constraint condition includes that each borrower obtains funds from at most S fund providers, and S is an integer greater than 1 and less than m. This means that the number of non-zero elements in each column of the matching result matrix is required to be less than or equal to S.

[0065] The consistency constraint condition includes that the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to a predetermined interest rate difference threshold, where the m weighted interest rates correspond one-to-one with the m fund providers.

[0066] The weighted interest rate corresponding to each fund provider is the weighted average of the borrowing interest rates of all borrowers who match the transaction amount with the fund provider, and the weight of the weighted average is the ratio of the transaction amount to the contribution amount of the fund provider.

[0067] For convenience, denote the contribution amounts of the m fund providers as vector , denote the borrowing amounts of the n borrowers as vector , denote the borrowing interest rates of the n borrowers as vector , and the m weighted interest rates can form a column vector , and its quantitative expression can be calculated by the following formula (1):

[0068] (1)

[0069] In the above formula, H represents the matching result matrix, represents element-wise division.

[0070] According to an embodiment of the present disclosure, the fund supply information and the borrowing demand information can be matched based on the sparsity constraint condition and the consistency constraint condition in an operational research optimization manner to obtain a matching result matrix.

[0071] Specifically, an operational research optimization model can be constructed according to the sparsity constraint condition and the consistency constraint condition, and the above-mentioned matching result matrix can be obtained by solving using this operational research optimization model.

[0072] In one embodiment, the objective function of the operational research optimization model can be as shown in the following formula (2):

[0073] Min (2)

[0074] Wherein, is the total contribution of the i-th contributor, is the borrowing amount of the j-th borrower.

[0075] Correspondingly, denote as the -th column of

[0076] (3)

[0077] The consistency constraint condition can be as shown in the following formula (4):

[0078] (4)

[0079] In formula (4), is the interest rate difference threshold (such as 2 basis points, where one basis point is equal to 0.01%, that is, one hundredth of 1%).

[0080] In one embodiment, when constructing the objective function of the operational research optimization model, the objective function shown in formula (2) can be optimized according to the sparsity constraint condition and / or the consistency constraint condition.

[0081] For example, according to the sparsity constraint condition, a sparsity penalty function (Smoothly Clipped Absolute Deviation, abbreviated as SCAD) and / or a consistency penalty function can be added to the objective function shown in Equation (2). Then, by minimizing the sparsity penalty function SCAD and / or the consistency penalty function, a matching result matrix is obtained. In this case, according to requirements such as the computational amount during the solution process or the focus of the solution, only one of the sparsity penalty function SCAD or the consistency penalty function can be selected to be added to the objective function, or the sparsity penalty function SCAD or the consistency penalty function can be added simultaneously according to a certain weight ratio. This can adjust and optimize the solution result more flexibly, and obtain a more optimized or suitable matching result matrix.

[0082] Finally, in operation S204, based on the transaction amount in the matching result matrix, a fund transaction is carried out between m fund providers and n borrowers.

[0083] It can be seen that the embodiments of the present disclosure can overcome the problems of low efficiency and difficulty in obtaining an optimized solution through manual processing in the prior art when transaction matching is required. The fund supply information and borrowing demand information can be matched based on the sparsity constraint condition and the consistency constraint condition in the manner of operations research optimization to obtain a matching result matrix, and then transactions are carried out according to the trading volume in the matching result matrix, which can achieve efficient matching, and the matching result is more optimized than the manual solution.

[0084] Figure 3 The flowchart of the transaction matching method according to another embodiment of the present disclosure is schematically shown.

[0085] As Figure 3 shown, the transaction matching method according to this embodiment may include operation S201, operation S202, operations S213 to S253, and operation S204.

[0086] First, in operation S201, fund supply information is obtained, where the fund supply information includes m fund providers and the investment amount of each fund provider, and m is an integer greater than 1.

[0087] And in operation S202, borrowing demand information is obtained, where the borrowing demand information includes n borrowers and the borrowing amount and borrowing interest rate of each borrower, and n is an integer greater than 1.

[0088] Next, in operation S213, without considering the consistency constraint condition, the fund supply information and the borrowing demand information are matched based on the sparsity constraint condition to obtain a first matching matrix.

[0089] For example, a sparsity solver can be used to match the fund supply information and borrowing demand information to obtain a first matching matrix. The sparsity solver is an operations research optimization model, and the objective function of the sparsity solver includes minimizing a sparse penalty function. For example, a sparse penalty function SCAD can be added to the objective function shown in Equation (2), and then the first matching matrix can be obtained by minimizing the sparse penalty function SCAD. .

[0090] Then, in operation S223, m weighted interest rates are calculated based on the first matching matrix.

[0091] In operation S233, it is determined whether the difference between the maximum value and the minimum value among the calculated m weighted interest rates is less than or equal to the interest rate difference threshold.

[0092] If the judgment result of operation S233 is yes, operation S253 is executed to determine the first matching matrix as the matching result matrix. And in operation S204, according to the transaction amount in the matching result matrix determined in operation S253, fund transactions are carried out between the m fund providers and the n borrowers.

[0093] Since the consistency constraint condition is not considered during the matching in operation S213, after obtaining the first matching matrix, it is determined whether the transaction plan based on the first matching matrix can meet the consistency constraint condition. If the first matching matrix meets the consistency constraint condition, the first matching matrix is directly output . In this way, the matching result matrix can be quickly obtained, improving the efficiency of transaction matching and transaction execution.

[0094] If the judgment result of operation S233 is no, operations S243, S223, and S233 are looped until the judgment result of operation S233 is that the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold, at which point the loop ends.

[0095] Specifically, in operation S243, the first matching matrix is optimized. In operation S223, m weighted interest rates are recalculated based on the optimized and updated first matching matrix. Then, in operation S233, it is re-determined whether the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold. When the judgment result of operation S233 is that the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold, it indicates that the optimized first matching matrix can meet the consistency constraint condition, and the loop ends.

[0096] Among them, the process of optimizing the first matching matrix in operation S243 is as Figure 4 shown.

[0097] Figure 4 The flowchart shows the optimization of the first matrix in operation S243 of the transaction matching method according to another embodiment of the present disclosure.

[0098] As Figure 4 shown, operation S243 may include operations S401 to S404.

[0099] In operation S401, determine the first fund provider corresponding to the maximum value among the m weighted interest rates and the second fund provider corresponding to the minimum value.

[0100] In operation S402, obtain local fund supply information and local fund demand information.

[0101] Specifically, the local fund supply information includes the investment amounts of the first fund provider and the second fund provider. In one embodiment, after determining the first fund provider and the second fund provider, the local fund attack learning can be extracted from the fund supply information.

[0102] The local fund demand information includes all borrowers to be optimized and the local borrowing amount of each borrower to be optimized. Among them, the borrower to be optimized is the borrower who matches the transaction amount with at least one of the first fund provider and the second fund provider in the first matching matrix, and the local borrowing amount is the sum of the transaction amounts that the borrower to be optimized matches with at least one of the first fund provider and the second fund provider in the first matching matrix.

[0103] In one embodiment, the rows corresponding to the first fund provider and the second fund provider can be first extracted from the first matching matrix to form a first low-dimensional matrix; then the columns with all zero elements are removed from the first low-dimensional matrix to obtain a second low-dimensional matrix; then the local fund demand information and the local fund demand information are extracted from the second low-dimensional matrix. Among them, the local fund supply information can be obtained according to the fund provider corresponding to the row in the second low-dimensional matrix and the sum of the row elements, and the local fund demand information can be obtained according to the borrower corresponding to the column in the second low-dimensional matrix and the sum of the column elements. In this way, the local fund supply information and the local fund demand information can be quickly obtained by processing the first matching matrix.

[0104] In operation S403, match the local fund supply information and the local fund demand information based on the local consistency condition to obtain a second matching matrix.

[0105] It can be seen that the number of rows of the second matching matrix is 2, and each element represents the transaction amount between one of the first fund provider and the second fund provider and a borrower to be optimized. Among them, the local consistency condition includes that the difference between the weighted interest rate corresponding to the first fund provider and the weighted interest rate corresponding to the second fund provider is less than or equal to the interest rate difference threshold.

[0106] Then, in operation S404, the transaction amounts of the first fund provider and the second fund provider in the first matching matrix are correspondingly replaced by using the transaction amounts in the second matching matrix to update the first matching matrix.

[0107] In one embodiment, a consistency solver may be used to match the local fund supply information and the local fund demand information, and the second matching matrix is obtained by solving. The consistency solver is an operations research optimization model. The objective function of this consistency solver includes minimizing a consistency penalty function. For example, a consistency penalty function may be appended to the objective function shown in Equation (2), and the second matching matrix is obtained by minimizing this consistency penalty function.

[0108] Among them, the consistency penalty function is a function obtained based on the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider.

[0109] In one embodiment, this consistency penalty function may be as shown in the following Equation (5):

[0110] (5)

[0111] In Equation (5), 、 respectively represent the weighted interest rates of the first fund provider and the second fund provider.

[0112] In one embodiment, the objective function of this consistency solver may include minimizing the weighted value of the consistency penalty function and the sparse penalty function. By adjusting the weights of the consistency penalty function and the sparse penalty function, the second matching matrix can reduce the difference between the weighted interest rates of the first fund provider and the second fund provider while obtaining a solution that is as sparse as possible, and minimize the impact of this tuning process on the sparsity of the first matching matrix.

[0113] In the embodiments of the present disclosure, when the first matching matrix does not meet the consistency constraint conditions, the local fund supply information and the local fund demand information are matched based on the local consistency conditions, and the second matching matrix is solved. This process is a low-dimensional local adjustment. Since the solution complexity of non-linear optimization is proportional to the cube of the number of variables, this will greatly accelerate the solution process and greatly improve the transaction matching efficiency.

[0114] Figure 5 Schematically shows a flowchart of a transaction matching method according to still another embodiment of the present disclosure.

[0115] As Figure 5As shown, the transaction matching method according to this embodiment may include steps 001 to 004. This method can be executed by a transaction matching system, where the transaction matching system can integrate a data processing module and a solving module to automatically process the data of the business system and quickly output a transaction matching result that meets the sparsity constraint condition and the consistency constraint condition. The specific content of each step is as follows. Among them, step 001 is executed by the data processing module, and the remaining part is executed by the solving module. First, the data processing module reads the data of the lender and the borrower from the transaction database. After some basic database processing work, the required data can be integrated into a data frame and passed to the solving module, and the solving module solves the matching result matrix H. The specific process is introduced as follows.

[0116] First, in step 001, data acquisition. First, the data processing module reads the fund supply information and the borrowing demand information from the database of the business system and processes them into the input data of the solving module. Specifically, it includes reading the relevant data of the fund providers and the borrowers. After screening by the database software, the data can be integrated into a data frame, which contains 3 columns of data in total. Among them, the first column of data corresponds to the investment amount of each fund provider; the second column of data corresponds to the borrowing amount of each borrower; the third column corresponds to the borrowing interest rate promised by each borrower (the order of the borrowers in the third column of data is the same as that in the second column).

[0117] Suppose there are fund providers and borrowers. Then the first column of data is a column vector with a length of , while the second and third columns of data are column vectors with a length of , and the order of the borrowers corresponding to the two is the same. After obtaining the processed data frame, the data frame is passed to the solving module.

[0118] Then, in step 002, in the solving module, in the way of operations research optimization, the problem of transaction matching according to the fund supply information and the borrowing demand information is modeled into a high-dimensional sparse optimization problem with constraints.

[0119] As the assumption in step 001, if there are m fund providers and n borrowers, the result of the transaction matching can be modeled as a matrix , where the element represents the transaction amount between the i-th fund provider and the j-th borrower.

[0120] The so-called matching the fund supply information and the borrowing demand information based on the sparsity constraint condition and the consistency constraint condition is to find a matrix , satisfying the sparsity constraint condition and the consistency constraint condition, and having a relatively fast solution speed at the same time. Among them, in the sparsity constraint condition, it is required that the orders of a single borrower should not be split too scattered. In fact, it is to ensure that the non-zero elements in each column of the matrix are fewer. In this way, under the condition of satisfying the consistency constraint condition, by minimizing the sparse penalty function SCAD, the transaction matching problem including the constraint condition can be modeled into a constrained high-dimensional sparse optimization problem.

[0121] Next, in step 003, call the sparsity solver to solve. For the optimization problem modeled in the previous step, temporarily remove the consistency constraint condition, and then use the sparsity solver to obtain the first matching matrix . The sparsity solver can solve the optimization problem that contains the sparsity constraint condition but removes the consistency constraint condition by using a non-linear optimizer. By removing the complex consistency condition constraints, the solution efficiency can be greatly improved.

[0122] The solution obtained in this way, that is, the first matching matrix Of course, it cannot be guaranteed to satisfy the strict consistency constraint condition. This is because the sparsity solver simply does not consider the consistency constraint condition. Therefore, when obtaining the first matching matrix , it is necessary to make a judgment according to the consistency constraint condition. If the first matching matrix satisfies the consistency constraint condition, directly output the result; if not, go to the next step and use the consistency solver to continue the optimization and solution.

[0123] In step 004, call the consistency solver to solve. Note that if the solution obtained in step 003 (that is, the first matching matrix ) already satisfies the consistency constraint condition, there is no need to use the consistency solver anymore, otherwise the consistency solver is called to solve.

[0124] Specifically, using the consistency solver is to continuously optimize the first matching matrix through iteration.

[0125] Specifically, in each round of iteration, first find the current first matching matrix Among them, find the contribution amounts of the two fund providers with the largest and smallest weighted interest rates, that is, the contribution amounts of a pair of fund providers with the most disparate weighted interest rates. For these 2 fund providers and the borrowers with which they have currently reached a transaction amount, model them into a low-dimensional constrained sparse optimization problem in a similar way to step 002. And use the solution result of the consistency solver to replace the transaction amounts of the corresponding fund providers and borrowers in the first matching matrix , and in this way, for the first matching matrix Perform local optimization. Note that the consistency solver is a local solver, and only the sparse optimization problem in the low-dimensional case needs to be solved for local fine-tuning. Since the solution complexity of nonlinear optimization is proportional to the cube of the number of variables, this will greatly accelerate the solution process. Using the consistency solver, local adjustments can be made step by step to narrow the gap in the weighted interest rates of the investors.

[0126] Repeat the iteration and continuously update the first matching matrix and continuously narrow the first matching matrix until the most significant interest rate gap in it meets the consistency constraint conditions. It can be mathematically proven that the consistency solver will necessarily make the transaction matching result meet the consistency constraint conditions while satisfying the sparsity constraint conditions as much as possible.

[0127] It can be seen that since the transaction matching problem can be modeled as a constrained high-dimensional sparse optimization problem, even with consistency constraint conditions and sparsity constraint conditions, it can still be directly solved by a nonlinear optimizer. Further, considering directly using a nonlinear optimizer to solve the original problem, when the number of variables m×n to be solved is large and the technical constraints are complex, the efficiency may be too low, resulting in too low transaction matching efficiency. A heuristic method can be adopted in the solution module, and two auxiliary nonlinear solvers, namely the sparsity solver and the consistency solver, are arranged in a hierarchical manner and solved in two steps to improve the transaction matching efficiency.

[0128] It can be seen that in the embodiments of the present disclosure, the data processing module transaction component system automatically extracts relevant data from the business system, and after preparation, inputs the standardized data into the solution module. The solution module innovatively combines the heuristic algorithm and the nonlinear optimizer, and can efficiently output the transaction matching result while satisfying the technical constraints.

[0129] The embodiments of the present disclosure can integrate the data processing module and the solution module through the transaction matching system, automatically process the data of the business system, and perform automatic matching of the fund supply information and the borrowing demand information, and can solve and give the transaction matching result that meets all technical constraints at a relatively fast speed.

[0130] In addition, in step 004 above, in order to reduce the impact on the sparsity of the result when the consistency solver is called to locally optimize the first matching matrix, the objective function of the consistency solver can incorporate consistency penalty and sparsity penalty. In one embodiment, the objective function of the consistency solver can be constructed in the manner shown by Figure 6 .

[0131] Figure 6 Schematically shows the flowchart of constructing a consistency solver in the transaction matching method of another embodiment of the present disclosure.

[0132] As Figure 6As shown, in one embodiment of the present disclosure, the process of constructing a consistency solver may include operations S601 to S606.

[0133] First, in operation S601, set the objective function of the consistency solver. The objective function of the consistency solver includes minimizing the weighted value of the consistency penalty function and the sparse penalty function. Set the weight of the consistency penalty function to 1 and set the initial value of the weight of the sparse penalty function.

[0134] Next, in operation S602, obtain the solution matrix obtained by the consistency solver matching the local fund supply information and the local fund demand information.

[0135] In operation S603, compare the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider obtained based on the solution matrix with the interest rate difference threshold to obtain a comparison result.

[0136] In operation S604, determine whether the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is less than or equal to the interest rate difference threshold. If not, execute operation S605; if so, execute operation S606.

[0137] In operation S605, when the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is greater than the interest rate difference threshold, lower the weight of the sparse penalty function in the objective function of the consistency solver.

[0138] In operation S606, output the objective function of the consistency solver.

[0139] For example, briefly denote the objective function of the consistency solver as , as shown in the following formula (6)

[0140] (6)

[0141] Where is the consistency penalty function, which can be as shown in the foregoing formula (5), is the determined sparse penalty function SCAD, for example, it can be an L1 regularization (Lasso regularization) function.

[0142] And, The functions themselves do not need to adjust parameters. Therefore, the only thing to be determined in the objective function shown in formula (6) is the sparse penalty weight parameter , and the choice of the sparse penalty weight parameter is crucial because the choice of this parameter will directly affect the solution effect. The permission parameter The larger it is, the more the objective function favors sparse penalty, and the sparser the solution; the parameter The smaller it is, the more the objective function favors consistency penalty, and the more the solution can satisfy the consistency constraint.

[0143] In the above operation S605, when reducing the weight parameter of the sparse penalty function in the objective function of the consistency solver it can be adjusted according to a certain step size (such as a fixed step size or a step size determined by a certain strategy) to avoid the unknown risks brought by artificially selecting this weight parameter

[0144] In one embodiment, the weight of the sparse penalty function in the objective function of the consistency solver can be reduced in the form of exponential decay. Specifically, since the solution result of the consistency solver is sensitive to the selection of the parameter In order to ensure the consistency technical constraints while not splitting the borrower's transactions too much as much as possible, the weight parameter can be set in the form of exponential decay. In this way, if the initial value of the weight parameter cannot narrow the gap between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider, the parameter is reduced by half, making the objective function more biased towards consistency penalty until the interest rate gap can be narrowed. In this way, there is no need to manually adjust the setting of the hyperparameter, and it can also adaptively ensure that the interest rate gap is narrowed each time. Then, through iteration, the transaction matching result can finally satisfy both the consistency constraint adjustment and the sparsity constraint conditions.

[0145] It can be seen that through the automated and efficient transaction matching system in the embodiments of the present disclosure, compared with the existing manual transaction matching methods, it can integrate business system data processing and automatic solution, and systematically improve the transaction matching efficiency. The solution obtained by the solution module can ensure that the sparsity constraint conditions and consistency constraint conditions required by the business are satisfied. And by arranging two auxiliary solution solvers in a ladder in the solution module, it eliminates the need for manual adjustment of hyperparameters (that is, the weight parameter ), improves the solution rate, and the overall solution efficiency meets the business requirements.

[0146] Figure 7 The block diagram of a transaction matching device 700 according to an embodiment of the present disclosure is schematically shown.

[0147] As Figure 7 shown, the transaction matching device 700 may include a first acquisition module 710, a second acquisition module 720, a matching module 730, and a transaction module 740. The transaction matching device 700 can execute the transaction matching method introduced with reference to Figures 2 - 6

[0148] ​​The first acquisition module 710 is configured to acquire fund supply information, where the fund supply information includes m fund providers and the investment amount of each fund provider, and m is an integer greater than 1. In one embodiment, the first acquisition module 710 may perform operation S201 introduced above.

[0149] The second acquisition module 720 is configured to acquire borrowing demand information, where the borrowing demand information includes n borrowers and the borrowing amount and borrowing interest rate of each borrower, and n is an integer greater than 1. In one embodiment, the second acquisition module 720 may perform operation S202 introduced above.

[0150] The matching module 730 is configured to match the fund supply information and the borrowing demand information by means of operations research optimization based on the sparsity constraint condition and the consistency constraint condition, and obtain a matching result matrix, where the matching result matrix is an m*n matrix, and the element in the i-th row and the j-th column represents the transaction amount between the i-th fund provider and the j-th borrower. Among them, the sparsity constraint condition includes that each borrower can obtain funds from at most S fund providers, and S is an integer greater than 1 and less than m. The consistency constraint condition includes that the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to a predetermined interest rate difference threshold, where the m weighted interest rates correspond one-to-one with the m fund providers, and the weighted interest rate corresponding to each fund provider is the weighted average of the borrowing interest rates of all borrowers who match the transaction amount with the fund provider, and the weight of the weighted average is the ratio of the transaction amount to the investment amount of the fund provider. In one embodiment, the matching module 730 may perform operation S203 introduced above.

[0151] The transaction module 740 is configured to conduct fund transactions between the m fund providers and the n borrowers based on the transaction amounts in the matching result matrix. In one embodiment, the transaction module 740 may perform operation S204 introduced above.

[0152] Figure 8 The block diagram of the matching module 730 according to an embodiment of the present disclosure is schematically shown.

[0153] As Figure 8 shown, according to this embodiment, the matching module 730 may include a first matching sub-module 731 and a tuning sub-module 732.

[0154] The first matching sub-module 731 is configured to: match the fund supply information and the borrowing demand information based on the sparsity constraint condition, and obtain a first matching matrix. In one embodiment, the first matching sub-module 731 may perform operation S213 introduced above.

[0155] The tuning sub-module 732 is used for: first calculating m weighted interest rates based on the first matching matrix; then when the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold, determining the first matching matrix as the matching result matrix, or when the difference between the maximum value and the minimum value among the m weighted interest rates is greater than the interest rate difference threshold, looping to perform the following operations to tune the first matching matrix until the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold. In one embodiment, the tuning sub-module 732 may perform the operation S243 introduced above.

[0156] Specifically, the tuning sub-module 732 may be used to loop and perform the operations S401 to S404 introduced above when the difference between the maximum value and the minimum value among the m weighted interest rates is greater than the interest rate difference threshold until the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold.

[0157] In one embodiment, the tuning sub-module includes a consistency solver. Among them, the tuning sub-module 732 is further used for: using the consistency solver to match the local fund supply information and the local fund demand information, and solving to obtain a second matching matrix. Among them, the objective function of the consistency solver includes minimizing the weighted value of the consistency penalty function and the sparse penalty function. The consistency penalty function is a function obtained based on the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider.

[0158] In one embodiment, the tuning sub-module 732 further includes a consistency solver construction sub-module. The consistency solver construction module is used to construct a consistency solver, including setting the weight of the consistency penalty function in the objective function of the consistency solver to 1, and setting the weight of the sparse penalty function to be iterated in the following manner after initialization: obtaining the solution matrix obtained by the consistency solver matching the local fund supply information and the local fund demand information; comparing the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider obtained based on the solution matrix with the interest rate difference threshold to obtain a comparison result; when the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is greater than the interest rate difference threshold, reducing the weight of the sparse penalty function in the objective function of the consistency solver to update the objective function of the consistency solver, and using the updated consistency solver to re-match the local fund supply information and the local fund demand information.

[0159] In one embodiment, the consistency solver construction sub-module is specifically used for: when the difference between the weighted interest rate of the first fund provider and the weighted interest rate of the second fund provider in the comparison result is greater than the interest rate difference threshold, reducing the weight of the sparse penalty function in the objective function of the consistency solver in the form of exponential decay.

[0160] In one embodiment, the first matching sub-module 731 further includes a sparsity solver, and the first matching sub-module is further configured to: match the fund supply information and the borrowing demand information by using the sparsity solver, and solve to obtain a first matching matrix; wherein, the objective function of the sparsity solver includes minimizing a sparse penalty function.

[0161] According to an embodiment of the present disclosure, any plurality of modules among the first acquisition module 710, the second acquisition module 720, the matching module 730, the transaction module 740, the first matching sub-module 731, the tuning sub-module 732, the data processing module, and the solving module can be combined and implemented in one module, or any one of them can be split into multiple modules. Or, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the first acquisition module 710, the second acquisition module 720, the matching module 730, the transaction module 740, the first matching sub-module 731, the tuning sub-module 732, the data processing module, and the solving module can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by any other reasonable means such as hardware or firmware by integrating or packaging circuits, or can be implemented in any one of the three implementation manners of software, hardware, and firmware or in any suitable combination of several of them. Or, at least one of the first acquisition module 710, the second acquisition module 720, the matching module 730, the transaction module 740, the first matching sub-module 731, the tuning sub-module 732, the data processing module, and the solving module can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding functions can be executed.

[0162] Figure 9 The structural block diagram of an electronic device suitable for implementing the transaction matching method of the embodiments of the present disclosure is schematically shown.

[0163] As Figure 9As shown, the electronic device 900 according to an embodiment of the present disclosure includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage section 908 into a random access memory (RAM) 903. The processor 901 can include, for example, a general microprocessor (such as a CPU), an instruction set processor, and / or a related chipset, and / or a dedicated microprocessor (such as, an application specific integrated circuit (ASIC)), etc. The processor 901 can also include on-board memory for caching purposes. The processor 901 can include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0164] In the RAM 903, various programs and data required for the operation of the electronic device 900 are stored. The processor 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. The processor 901 performs various operations of the method flow according to an embodiment of the present disclosure by executing the program in the ROM 902 and / or the RAM 903. It should be noted that the program can also be stored in one or more memories other than the ROM 902 and the RAM 903. The processor 901 can also perform various operations of the method flow according to an embodiment of the present disclosure by executing the program stored in the one or more memories.

[0165] According to an embodiment of the present disclosure, the electronic device 900 can further include an input / output (I / O) interface 905, and the input / output (I / O) interface 905 is also connected to the bus 904. The electronic device 900 can further include one or more of the following components connected to the I / O interface 905: an input portion 906 including a keyboard, a mouse, etc.; an output portion 907 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage portion 908 including a hard disk, etc.; and a communication portion 909 including a network interface card such as a LAN card, a modem, etc. The communication portion 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as needed. A removable medium 911, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 910 as needed so that a computer program read therefrom can be installed into the storage portion 908 as needed.

[0166] The present disclosure also provides a computer-readable storage medium, which can be included in the device / apparatus / system described in the above embodiments; or can exist separately without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the one or more programs are executed, the method according to an embodiment of the present disclosure is implemented.

[0167] According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: portable computer disks, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), portable compact disk read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the above. In the present disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present disclosure, the computer-readable storage medium may include the above-described ROM 902 and / or RAM 903 and / or one or more memories other than ROM 902 and RAM 903.

[0168] An embodiment of the present disclosure also includes a computer program product, which includes a computer program that contains program code for executing the method shown in the flowchart. When the computer program product runs in a computer system, the program code is used to enable the computer system to implement the method provided by the embodiment of the present disclosure.

[0169] When the computer program is executed by the processor 901, it executes the above functions defined in the system / apparatus of the embodiment of the present disclosure. According to an embodiment of the present disclosure, the above-described system, apparatus, module, unit, etc. can be implemented by computer program modules.

[0170] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices and magnetic storage devices. In another embodiment, the computer program may also be transmitted and distributed in the form of a signal on a network medium, and be downloaded and installed through the communication part 909, and / or be installed from the removable medium 911. The program code included in the computer program can be transmitted by any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0171] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 909, and / or be installed from the removable medium 911. When the computer program is executed by the processor 901, it executes the above functions defined in the system of the embodiment of the present disclosure. According to an embodiment of the present disclosure, the above-described system, device, apparatus, module, unit, etc. can be implemented by computer program modules.

[0172] According to embodiments of the present disclosure, program code for executing the computer programs provided by the embodiments of the present disclosure may be written in any combination of one or more programming languages. Specifically, these computing programs may be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. The programming languages include, but are not limited to, programming languages such as Java, C++, Python, the "C" language, or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., by connecting through the Internet using an Internet service provider).

[0173] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and combinations of blocks in the block diagram or flowchart, may be implemented by a dedicated hardware-based system that performs the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0174] Those skilled in the art can understand that the features recited in the various embodiments and / or claims of the present disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly recited in the present disclosure. In particular, without departing from the spirit and teachings of the present disclosure, the features recited in the various embodiments and / or claims of the present disclosure can be combined and combined in various ways. All such combinations and / or combinations fall within the scope of the present disclosure.

[0175] The embodiments of the present disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments have been described separately above, this does not mean that the measures in each embodiment cannot be used advantageously in combination. The scope of the present disclosure is defined by the appended claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art can make various substitutions and modifications, and all such substitutions and modifications should fall within the scope of the present disclosure.

Claims

1. A transaction matching method, wherein: The method comprises: Acquire capital supply information, wherein the capital supply information includes m investors and the capital contribution amount of each investor, where m is an integer greater than 1; Obtaining loan demand information, wherein the loan demand information includes n borrowers and the loan amount and loan interest rate of each borrower, where n is an integer greater than 1; Based on the sparsity constraint and the consistency constraint, the fund supply information and the borrowing demand information are matched by means of operations optimization to obtain a matching result matrix, wherein the matching result matrix is ​​an m*n matrix, wherein the element in the i-th row and the j-th column represents the transaction amount between the i-th investor and the j-th borrower; and Based on the transaction amount in the matching result matrix, performing fund transactions between the m investors and the n borrowers; The sparsity constraint condition includes that each borrower can obtain funds from at most S investors, where S is an integer greater than 1 and less than m; The consistency constraint condition includes that the difference between the maximum and minimum values ​​of the m weighted interest rates is less than or equal to a predetermined interest rate difference threshold, wherein the m weighted interest rates correspond one-to-one to the m investors, and the weighted interest rate corresponding to each investor is the weighted average of the borrowing interest rates of all borrowers who match the transaction amount with the investor, and the weight of the weighted average is the proportion of the transaction amount to the investor's investment amount.

2. The method according to claim 1, wherein: Based on the sparsity constraint condition and the consistency constraint condition, the fund supply information and the borrowing demand information are matched by means of operations optimization to obtain a matching result matrix including: Matching the fund supply information and the borrowing demand information based on the sparsity constraint condition to obtain a first matching matrix; Calculating the m weighted rates based on the first matching matrix; When the difference between the maximum value and the minimum value among the m weighted interest rates is less than or equal to the interest rate difference threshold, the first matching matrix is ​​determined to be the matching result matrix.

3. The method according to claim 2, wherein: The method further includes, when the difference between the maximum value and the minimum value of the m weighted interest rates is greater than the interest rate difference threshold, loopingly performing the following operations until the difference between the maximum value and the minimum value of the m weighted interest rates is less than or equal to the interest rate difference threshold: Determine a first investor corresponding to the maximum value and a second investor corresponding to the minimum value among the m weighted interest rates; Acquire local fund supply information and local fund demand information; wherein the local fund supply information includes the fund contribution of the first investor and the fund contribution of the second investor; the local fund demand information includes all borrowers to be optimized and the local loan amount of each borrower to be optimized, wherein the borrower to be optimized is a borrower that matches a transaction amount with at least one of the first investor and the second investor in the first matching matrix, and the local loan amount is the sum of the transaction amounts matched between the borrower to be optimized and at least one of the first investor and the second investor in the first matching matrix; Matching the local fund supply information and the local fund demand information based on a local consistency condition to obtain a second matching matrix, wherein the number of rows of the second matching matrix is ​​2, and each element represents the transaction amount between one of the first investor and the second investor and one of the borrowers to be optimized; Using the transaction amounts in the second matching matrix, correspondingly replacing the transaction amounts of the first investor and the second investor in the first matching matrix, so as to update the first matching matrix; Recalculating the m weighted rates based on the updated first matching matrix to update the m weighted rates; Comparing the difference between the maximum and minimum values ​​of the updated m weighted interest rates with the interest rate difference threshold; The local consistency condition includes that the difference between the weighted interest rate corresponding to the first investor and the weighted interest rate corresponding to the second investor is less than or equal to the interest rate difference threshold.

4. The method according to claim 3, wherein: The obtaining of local fund supply information and local fund demand information includes: Extracting rows corresponding to the first investor and the second investor from the first matching matrix to form a first low-dimensional matrix; Eliminate columns whose column elements are all zero from the first low-dimensional matrix to obtain a second low-dimensional matrix; and The local capital demand information and the local capital demand information are extracted from the second low-dimensional matrix.

5. The method according to claim 3, wherein: The matching of the local fund supply information and the local fund demand information based on the local consistency condition to obtain a second matching matrix includes: Matching the local fund supply information and the local fund demand information using a consistency solver to obtain the second matching matrix; The objective function of the consistency solver includes minimizing the weighted value of the consistency penalty function and the sparse penalty function, wherein, The consistency penalty function is a function obtained based on the difference between the weighted interest rate of the first investor and the weighted interest rate of the second investor.

6. The method according to claim 5, wherein: The weight of the consistency penalty function in the objective function of the consistency solver is 1, and the weight of the sparse penalty function is iteratively determined in the following way after setting the initial value: Obtaining a solution matrix obtained by matching the local fund supply information and the local fund demand information with the consistency solver; Comparing the difference between the weighted interest rate of the first investor and the weighted interest rate of the second investor obtained based on the solution matrix with the interest rate difference threshold to obtain a comparison result; When the difference between the weighted interest rate of the first investor and the weighted interest rate of the second investor in the comparison result is greater than the interest rate difference threshold, the weight of the sparse penalty function in the objective function of the consistency solver is lowered to update the objective function of the consistency solver, and the updated consistency solver is used to re-match the local funds supply information and the local funds demand information.

7. The method according to claim 6, wherein: The step of lowering the weight of the sparse penalty function in the objective function of the consistency solver includes lowering the weight of the sparse penalty function in the objective function of the consistency solver in the form of exponential decay.

8. The method according to claim 2, wherein: The matching of the fund supply information and the borrowing demand information based on the sparsity constraint condition to obtain a first matching matrix includes: Matching the fund supply information and the borrowing demand information using a sparsity solver to obtain the first matching matrix; Wherein, the objective function of the sparsity solver includes minimizing the sparsity penalty function.

9. A transaction matching device, wherein: The device comprises: A first acquisition module is used to acquire capital supply information, wherein the capital supply information includes m investors and the capital contribution of each investor, where m is an integer greater than 1; The second acquisition module is used to acquire loan demand information, wherein the loan demand information includes n borrowers and the loan amount and loan interest rate of each borrower; n is an integer greater than 1; A matching module, for matching the fund supply information and the borrowing demand information by means of operations optimization based on a sparsity constraint and a consistency constraint, to obtain a matching result matrix, wherein the matching result matrix is ​​an m*n matrix, wherein the elements in the i-th row and the j-th column represent the transaction amount between the i-th investor and the j-th borrower; wherein the sparsity constraint includes that each borrower obtains funds from at most S investors, S is an integer greater than 1 and less than m; the consistency constraint includes that the difference between the maximum and minimum values ​​of the m weighted interest rates is less than or equal to a predetermined interest rate difference threshold, wherein the m weighted interest rates correspond to the m investors one-to-one, and the weighted interest rate corresponding to each investor is the weighted average of the borrowing interest rates of all borrowers that match the transaction amount with the investor, and the weight of the weighted average is the proportion of the transaction amount to the investment amount of the investor; and A transaction module is used to perform fund transactions between the m investors and the n borrowers based on the transaction amount in the matching result matrix.

10. An electronic device comprising: one or more processors; a memory for storing one or more computer programs, The one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program or instruction stored thereon, wherein: When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

12. A computer program product comprising a computer program or instructions, wherein: When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images