Transaction processing method and device, equipment and storage medium

Abstract

The embodiment of the invention discloses a transaction processing method and device, equipment and a storage medium, and can be applied to the fields of computer technologies, block chains and the like. The method comprises the following steps: acquiring payment information reported by a transaction initiator in a process of requesting to acquire a payment link of a current transaction; determining account abnormity evaluation features and transaction abnormity evaluation features of the current transaction based on the payment information; determining an account anomaly coefficient of the current transaction based on the account anomaly evaluation feature, and determining a transaction anomaly coefficient of the current transaction based on the transaction anomaly evaluation feature; and determining a transaction risk level of the current transaction based on the account abnormal coefficient and the transaction abnormal coefficient, and issuing a payment link based on the transaction risk level. By adopting the embodiment of the invention, the transaction risk level of the current transaction can be effectively determined, the current transaction is processed according to the transaction risk level, and the applicability is high.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a transaction processing method, apparatus, device and storage medium. Background Art

[0002] The development of information technology has brought great convenience to people's lives. For example, the quick payment function based on information technology can enable online payment transactions between payers and payees.

[0003] However, in order to ensure the security of online transactions and prevent malicious programs from participating in transactions, it is often necessary to determine in advance the transaction risk level of each transaction so as to process the online transactions based on the transaction risk level. However, existing technologies often cannot accurately determine the transaction risk level of a transaction, and thus cannot effectively eliminate potential transaction risks. Summary of the invention

[0004] The embodiments of the present application provide a transaction processing method, apparatus, device, and storage medium, which can effectively determine the transaction risk level of the current transaction and have high applicability.

[0005] On the one hand, an embodiment of the present application provides a transaction processing method, the method comprising:

[0006] Obtain the payment information reported by the transaction initiator when requesting the payment link for the current transaction;

[0007] Determine the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction based on the above payment information;

[0008] Determine the account abnormality coefficient of the current transaction based on the above account abnormality assessment characteristics, determine the transaction abnormality coefficient of the current transaction based on the above transaction abnormality assessment characteristics; determine the transaction risk level of the current transaction based on the above account abnormality coefficient and the above transaction abnormality coefficient;

[0009] The current transaction is processed based on the above transaction risk level.

[0010] On the other hand, an embodiment of the present application provides a transaction processing device, the device comprising:

[0011] The information acquisition module is used to obtain the payment information reported by the transaction initiator when requesting to obtain the payment link of the current transaction;

[0012] A feature determination module, used to determine the account abnormality assessment feature and the transaction abnormality assessment feature of the current transaction based on the above payment information;

[0013] A level determination module, used to determine the account abnormality coefficient of the current transaction based on the above-mentioned account abnormality assessment characteristics, determine the transaction abnormality coefficient of the current transaction based on the above-mentioned transaction abnormality assessment characteristics; determine the transaction risk level of the current transaction based on the above-mentioned account abnormality coefficient and the above-mentioned transaction abnormality coefficient;

[0014] The transaction processing module is used to process the current transaction based on the above transaction risk level.

[0015] Optionally, the above-mentioned account abnormality assessment feature includes statistical data of each information item associated with the current transaction corresponding to each first transaction statistical dimension in each time window;

[0016] The information items associated with the current transaction include at least one of the following:

[0017] Payer account number;

[0018] Recipient's account number;

[0019] Terminal information of the payment terminal;

[0020] IP information of the payment terminal;

[0021] Among them, each of the above time windows is continuous and located within a first preset time period before the current transaction initiation moment.

[0022] Optionally, the first transaction statistics dimension includes at least one of the following:

[0023] The total number of related-party transactions;

[0024] The total transaction amount of related-party transactions;

[0025] The total number of related transactions for which transaction subject verification is performed;

[0026] The total number of related-party transactions for which verification of the transaction subject has not been carried out;

[0027] The total number of related transactions verified by the transaction subject;

[0028] The total number of related transactions that failed to pass the verification of the transaction entity.

[0029] Optionally, the transaction abnormality assessment feature includes at least one of the following:

[0030] Statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period;

[0031] The payment environment change characteristics of the current transaction, which are used to characterize the change status of the payment environment of the current transaction;

[0032] Each information item and information item combination associated with the current transaction corresponds to statistical data of each third transaction statistical dimension, and each of the above information item combinations includes any two of the above information items;

[0033] The transaction abnormality indicator corresponding to at least one of the payer or payee of the current transaction.

[0034] Optionally, the second transaction statistics dimension includes at least one of the following:

[0035] The total number of related-party transactions;

[0036] The total transaction amount of related-party transactions;

[0037] The total number of associated transactions with successful payments;

[0038] The total transaction amount of the associated transactions that have been successfully paid;

[0039] The transaction time of the most recent related transaction from the time the current transaction was initiated;

[0040] The transaction amount of the most recent related transaction before the current transaction is initiated;

[0041] The third transaction statistics dimension includes at least one of the following:

[0042] The total number of related transactions for which transaction subject verification is performed;

[0043] The total number of related transactions verified by the transaction subject;

[0044] Trading frequency.

[0045] Optionally, the payment environment change feature includes at least one of the following:

[0046] first information used to indicate whether the terminal information of the payment terminal changes during the process of requesting to obtain the payment link;

[0047] Second information used to indicate whether the IP information of the payment terminal changes during the process of requesting to obtain the payment link;

[0048] Third information used to indicate whether the browser user agent changes during the process of requesting to obtain the payment link;

[0049] The duration from the time the current transaction is initiated to the time the above payment information is reported.

[0050] Optionally, the transaction abnormality indicator corresponding to the payer includes at least one of the following:

[0051] The total number of IP information of the payer;

[0052] The total number of pre-set risk IP information in the payer's IP information;

[0053] The total number of recipients;

[0054] The total number of pre-set risk payees;

[0055] The transaction abnormality indicator corresponding to the payee includes at least one of the following:

[0056] The total number of IP information of the payer;

[0057] The total number of pre-set risk IP information in the payer's IP information;

[0058] The total number of payers;

[0059] The total number of pre-set risk payers.

[0060] Optionally, the level determination module is used to:

[0061] Determine the abnormal account risk weight and abnormal transaction risk weight;

[0062] Based on the account risk weight, the transaction risk weight, the account anomaly coefficient and the transaction anomaly coefficient, the transaction risk level of the current transaction is determined.

[0063] Optionally, the transaction processing module is used to:

[0064] In response to the transaction risk level being lower than a preset risk level, issuing a payment link to the transaction initiator;

[0065] In response to the transaction risk level being greater than or equal to the preset risk level, the transaction initiator is subject to transaction subject verification, and a payment link is issued to the transaction initiator after the transaction initiator passes the transaction subject verification.

[0066] Optionally, the account anomaly coefficient of the current transaction determined based on the account anomaly assessment feature is determined based on an account anomaly assessment model, and the transaction anomaly coefficient of the current transaction determined based on the transaction anomaly assessment feature is determined based on a transaction anomaly assessment model;

[0067] The above account anomaly assessment model and the above transaction anomaly assessment model are obtained by training the training module based on the following method:

[0068] Determine a training sample set, the training sample set including account abnormal sample evaluation features and transaction abnormal sample evaluation features of multiple historical transactions;

[0069] Input the abnormal account sample evaluation feature of each of the above historical transactions into the first initial model to obtain the abnormal account sample coefficient of the corresponding historical transaction, and input the abnormal transaction sample evaluation feature of each of the above historical transactions into the second initial model to obtain the abnormal transaction sample coefficient of the corresponding historical transaction;

[0070] Determine the actual transaction risk level of each of the above historical transactions, and determine the predicted transaction risk level of the corresponding historical transaction based on the account abnormal sample coefficient and the transaction abnormal sample coefficient of each of the above historical transactions; determine the total training loss based on the predicted transaction risk level and the actual transaction risk level of each of the above historical transactions;

[0071] The first initial model and the second initial model are iteratively trained based on the training sample set until the total training loss meets the training end condition, and the first initial model and the second initial model at the time of stopping the training are respectively determined as the account anomaly assessment model and the transaction anomaly assessment model.

[0072] On the other hand, an embodiment of the present application provides an electronic device, including a processor and a memory, wherein the processor and the memory are connected to each other;

[0073] The memory is used to store computer programs;

[0074] The above-mentioned processor is used to execute the transaction processing method provided in the embodiment of the present application when calling the above-mentioned computer program.

[0075] On the other hand, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the transaction processing method provided by the embodiment of the present application.

[0076] On the other hand, an embodiment of the present application provides a computer program product, which includes a computer program. When the above computer program is executed by a processor, it implements the transaction processing method provided by the embodiment of the present application.

[0077] In the embodiment of the present application, the account abnormality assessment feature and the transaction abnormality assessment feature of the current transaction can be determined respectively through the payment information reported by the transaction initiator in the process of obtaining the payment link. Based on the account abnormality coefficient determined by the account abnormality assessment feature and the transaction abnormality coefficient determined by the transaction abnormality assessment feature, the transaction risk level of the current transaction can be accurately determined, so as to effectively process the current transaction based on the transaction risk level to eliminate potential transaction risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0079] Figure 1 It is a schematic diagram of the network structure provided by the embodiment of the present application;

[0080] Figure 2 It is a flowchart of a transaction processing method provided in an embodiment of the present application;

[0081] Figure 3 It is a schematic diagram of a scenario for determining abnormal assessment features of an account provided in an embodiment of the present application;

[0082] Figure 4 This is a schematic diagram of a scenario for determining transaction abnormality assessment features provided by an embodiment of the present application;

[0083] Figure 5 This is another schematic diagram of a scenario for determining transaction abnormality assessment features provided by an embodiment of the present application;

[0084] Figure 6 This is another scenario diagram for determining transaction abnormality assessment features provided by the embodiment of the present application.

[0085] Figure 7 It is a schematic diagram of a scenario for determining an abnormal coefficient of an account provided in an embodiment of the present application;

[0086] Figure 8 It is a schematic diagram of a scenario of an abnormal coefficient of a transaction account provided in an embodiment of the present application;

[0087] Fig. 9 It is a flowchart of the model training method provided in the embodiment of the present application;

[0088] Fig.10 It is a schematic diagram of the process framework of the transaction processing method provided in the embodiment of the present application;

[0089] Fig.11 It is a structural diagram of a transaction processing device provided in an embodiment of the present application;

[0090] Fig.12 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0091] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0092] See also Figure 1 , Figure 1 Schematic diagram of the network structure provided by the embodiment of the present application. Figure 1 As shown, the transaction initiator can realize online transactions through the payment terminal 100 and the backend server 200 of the corresponding merchant or shopping platform.

[0093] Specifically, the transaction initiator may initiate a transaction to the backend server 200 through the payment terminal 100, and report payment information associated with the current transaction to the backend server 200 during the process of requesting the backend server 200 to obtain the payment link of the current transaction.

[0094] For the backend server 200, the backend server 200 can determine the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction based on the payment information reported by the payment terminal 100, and determine the account abnormality coefficient of the current transaction based on the account abnormality assessment characteristics of the current transaction, and determine the transaction abnormality coefficient of the current transaction based on the transaction abnormality assessment characteristics of the current transaction. Finally, the backend server 200 can determine the transaction risk level of the current transaction based on the account abnormality coefficient and transaction abnormality coefficient of the current transaction, and thus process the current transaction based on the transaction risk level of the current transaction.

[0095] The payment terminal 100 may be a personal computer, a laptop computer, a smart phone, a tablet computer, a smart watch, an intelligent voice interaction device, a smart home appliance, a vehicle-mounted terminal, a smart wearable device, etc., but is not limited thereto.

[0096] The backend server 200 may be a server in a cloud network, an independent physical server, or a server cluster or distributed system composed of multiple physical servers. The specific requirements may be determined based on the actual application scenario and are not limited here. A cloud server refers to a server that can provide basic cloud computing services such as cloud database, cloud computing, cloud functions, cloud storage, network services, cloud communications, intermediate services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0097] The transaction processing method provided in the embodiment of the present application can be applied to the payment link of the system and server with payment transaction function, such as H5 payment link, etc. Figure 2 , Figure 2 It is a flowchart of the transaction processing method provided in the embodiment of the present application. Figure 2 The transaction processing method shown may specifically include the following steps:

[0098] Step S21: Obtain payment information reported by the transaction initiator when requesting to obtain the payment link of the current transaction.

[0099] In some feasible implementations, the current transaction is a transaction initiated by the transaction initiator to the backend server of the merchant or shopping platform at the current moment, and the transaction initiator needs to request the backend server to obtain the payment link of the current transaction after submitting the current transaction.

[0100] The backend server may obtain payment information that is reported to the backend server by the transaction initiator during the process of requesting to obtain the payment link of the current transaction.

[0101] Among them, the payment information reported by the transaction initiator when requesting to obtain the payment link of the current transaction includes the timestamp involved in each payment link (such as order submission link, etc.) before the payment link is issued to the payment initiator, the payee account number, the payee account number, the terminal information of the payment terminal, the transaction amount, and the browser UA (browser_ua) information used by the payment initiator.

[0102] It should be noted that the payment information reported by the transaction initiator in each payment link during the process of requesting to obtain the payment link of the current transaction can be the same or different, which is determined by the corresponding payment link. For example, if the payment terminal of the transaction initiator changes during the process of requesting to obtain the payment link of the current transaction, the transaction initiator will report the terminal information of the original payment terminal and the terminal information of the changed payment terminal.

[0103] Step S22: Determine the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction based on the payment information.

[0104] In some feasible implementations, the account abnormality assessment characteristics of the current transaction are used to assess the account risks of the accounts involved in the current transaction (such as the payer account and the payee account). When determining the account abnormality assessment characteristics of the current transaction, the account abnormality assessment characteristics of the current transaction can be determined based on information associated with at least one of the payer account or the payee account in the acquired payment information.

[0105] For example, the account abnormality assessment characteristics of the current transaction may be determined based on the payer account number in the acquired payment information and the transaction data associated therewith, and / or the payee account number in the payment information and the transaction data associated therewith.

[0106] Optionally, the account abnormality assessment feature of the current transaction may include statistical data of each information item associated with the current transaction corresponding to each first transaction statistical dimension in each time window.

[0107] Among them, each time window is continuous and located in the first preset time period before the current transaction initiation time, and the first preset time period can be any time period before the current transaction initiation time. For example, every minute within one hour before the current transaction initiation time is a time window, or every hour within 24 hours before the current transaction initiation time is a time window, which can be determined based on the actual application scenario requirements and is not limited here.

[0108] The information item associated with the current transaction includes at least one of the payer account number, payee account number, terminal information of the payment terminal, and IP information of the payment terminal corresponding to the current transaction.

[0109] If, during the process of requesting to obtain the payment link of the current transaction, any one of the payer account, payee account, terminal information of the payment terminal, and IP information of the payment terminal corresponding to the current transaction changes, the account abnormality assessment feature of the current transaction includes the information item before the change and the information item after the change, such as the terminal information of the payment terminal before the change and the terminal information after the change.

[0110] In the embodiment of the present application, the first transaction statistical dimension may be at least one of the following:

[0111] The total number of related-party transactions;

[0112] The total transaction amount of related-party transactions;

[0113] The total number of related transactions for which transaction subject verification is performed;

[0114] The total number of related-party transactions for which verification of the transaction subject has not been carried out;

[0115] The total number of related transactions verified by the transaction subject;

[0116] The total number of related transactions that failed to pass the verification of the transaction entity.

[0117] That is to say, for any information item associated with the current transaction, the account abnormality assessment feature of the current transaction may include any one or more of the total transaction amount of associated transactions of the information item in any time window, the total number of associated transactions that have been verified by the transaction subject, the total number of associated transactions that have not been verified by the transaction subject, the total number of associated transactions that have passed the verification of the transaction subject, or the total number of associated transactions that have not passed the verification of the transaction subject.

[0118] Among them, for the payee account of the current transaction (for the convenience of description, referred to as the first account below), the associated transactions of the first account are transactions in which the payee account is the first account; for the payer account of the current transaction (for the convenience of description, referred to as the second account below), the associated transactions of the second account are transactions in which the payer account is the second account; for the terminal information of the payment terminal of the current transaction (for the convenience of description, referred to as the first payment terminal below), the associated transactions of the terminal information of the first payment terminal are transactions in which the payment terminal is the first terminal; for the IP information of the first terminal of the current transaction, the associated transactions of the IP information of the first terminal are transactions in which the IP information is the IP information of the first terminal.

[0119] Among them, when determining the abnormal assessment characteristics of the account of the current transaction, the first set, the second set and the third set can be determined first.

[0120] The first set includes each information item of the current transaction, the second set includes each time window, and the third set includes each first transaction statistical dimension.

[0121] It is further possible to determine the Cartesian product of the first set, the second set, and the third set to obtain multiple first feature dimensions, each of which is used to characterize a first transaction statistical dimension of an information item of the current transaction in a time window. Based on this, the statistical data corresponding to each first feature dimension can be determined from the transaction database, and then the statistical data of each first feature dimension can be determined as the account abnormality assessment feature of the current transaction.

[0122] As an example, see Figure 3 , Figure 3 Schematic diagram of a scenario for determining abnormal assessment features of an account provided in an embodiment of the present application. Figure 3 As shown, the information items associated with the current transaction (first set) include the payee account, the payee account, the terminal information of the payment terminal, and the IP information of the payment terminal. The time window (second set) includes time windows 1-n, and the first transaction statistical dimension (third set) includes the total number of associated transactions, the total transaction amount of associated transactions, the total number of associated transactions verified by the transaction subject, and the total number of associated transactions verified by the transaction subject.

[0123] Determine the Cartesian product of the first set, the second set, and the third set to obtain multiple first feature dimensions, such as the total number of associated transactions of the payer account in time window 1, the total number of associated transactions of the payee account in time window n, etc. At this time, the statistical data corresponding to each first feature dimension is the account abnormality assessment feature of the current transaction.

[0124] Among them, any associated transaction in the embodiments of the present application may be a transaction in which payment is successful, or a transaction in which payment is failed (such as a transaction that has not been verified by the transaction subject, etc.).

[0125] Among them, the above-mentioned transaction subject verification is used to verify whether the transaction initiator is a program subject or a non-program subject, that is, to verify whether the corresponding transaction is initiated by a program subject (such as a computer program, plug-in application, and script, etc.) or a non-program subject (such as a natural person).

[0126] In some feasible implementations, the transaction anomaly assessment feature of the current transaction is used to assess the risk of the current transaction itself. When determining the transaction anomaly assessment feature of the current transaction, the transaction anomaly assessment feature of the current transaction can be determined based on the information associated with the current transaction in the acquired payment information (such as the payee's account number, payment environment, payer's account number, etc.).

[0127] Optionally, the transaction abnormality assessment feature of the current transaction may include at least one of the following:

[0128] Statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period;

[0129] The payment environment changes of the current transaction;

[0130] Each information item and information item combination associated with the current transaction corresponds to statistical data of each third transaction statistical dimension;

[0131] The transaction abnormality indicator corresponding to at least one of the payer or payee of the current transaction.

[0132] The information item associated with the current transaction includes at least one of the payer account number, payee account number, terminal information of the payment terminal, and IP information of the payment terminal corresponding to the current transaction.

[0133] Among them, the second preset time period can be the day when the current transaction is initiated, or it can be weekly, daily, monthly or annual, etc., which can be determined based on the actual application scenario requirements and is not limited here.

[0134] Among them, the current transaction initiation time can be determined according to the timestamp in the payment information.

[0135] For example, the statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period may include the statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension on the day, and may also include the statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension on a weekly, monthly or annual basis.

[0136] In the embodiment of the present application, the second transaction statistics dimension may include at least one of the following:

[0137] The total number of related-party transactions;

[0138] The total transaction amount of related-party transactions;

[0139] The total number of associated transactions with successful payments;

[0140] The total transaction amount of the associated transactions that have been successfully paid;

[0141] The transaction time of the most recent related transaction from the time the current transaction was initiated;

[0142] The transaction amount of the most recent related transaction before the current transaction is initiated.

[0143] It should be noted that the above-mentioned second transaction statistics dimension is only an example. For example, it may also include the total number of related transactions with failed payments, the total transaction amount of related transactions with failed payments, etc. The specific details can be determined based on the actual application scenario requirements and are not limited here.

[0144] For example, the statistical data of each second transaction statistical dimension corresponding to each information item associated with the current transaction in each second preset time period may include the total number of associated transactions on the day, the total transaction amount of associated transactions on the day, the total number of associated transactions with successful payments on the day, the total amount of associated transactions with successful payments on the day, the transaction time and transaction amount of the most recent associated transaction on the day from the time the current transaction was initiated, etc.

[0145] See also Figure 4 , Figure 4 Schematic diagram of a scenario for determining transaction abnormality assessment features provided by an embodiment of the present application. Figure 4 As shown, when determining the total number of related transactions on the day, the total transaction amount of related transactions on the day, the total number of related transactions with successful payments on the day, the total amount of related transactions with successful payments on the day, the transaction time and transaction amount of the most recent related transaction from the time the current transaction was initiated, etc. for each information item associated with the current transaction, the first set and the fourth set can be determined first.

[0146] The first set includes each information item of the current transaction, that is, the first set includes the payer account number, the payee account number, the terminal information of the payment terminal and the IP information of the payment terminal of the current transaction.

[0147] The fourth set includes multiple fourth transaction statistical dimensions, and the fourth transaction statistical dimensions are determined by the second preset time period (the day) and each second transaction statistical dimension. That is, the fourth set includes the fourth transaction statistical dimensions such as the total number of related transactions on the day, the total transaction amount of related transactions on the day, the total number of related transactions with successful payments on the day, the total transaction amount of related transactions with successful payments on the day, the transaction time of the most recent related transaction from the time when the current transaction was initiated, and the transaction amount of the most recent related transaction from the time when the current transaction was initiated.

[0148] Further, the Cartesian product of the first set and the fourth set may be determined to obtain multiple second feature dimensions, each of which is used to characterize a fourth transaction statistical dimension corresponding to an information item of the current transaction. For example, the second feature dimension may be used to characterize the total number of associated transactions of the payer account of the current transaction on the same day, the total number of associated transactions of the payer account that successfully paid on the same day, and so on.

[0149] Based on this, the statistical data corresponding to each second characteristic dimension can be determined from the transaction database, and then the total number of related transactions on the day, the total transaction amount of related transactions on the day, the total number of related transactions with successful payments on the day, the total amount of related transactions with successful payments on the day, and the transaction time and transaction amount of the most recent related transaction on the day from the time the current transaction was initiated can be obtained based on the statistical data of each second characteristic dimension.

[0150] For another example, the statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period may also include the total number of associated transactions for each information item associated with the current transaction within a week (month or year), the total transaction amount of associated transactions within a week (month or year), the total number of associated transactions with successful payments within a week (month or year), and the total amount of associated transactions with successful payments within a week (month or year).

[0151] See also Figure 5 , Figure 5 is another schematic diagram of a scenario for determining transaction abnormality assessment features provided by an embodiment of the present application. Figure 5 As shown, when determining the total number of associated transactions, the total transaction amount of associated transactions, the total number of associated transactions with successful payments, and the total amount of associated transactions with successful payments for each information item associated with the current transaction within each second preset time period (week, month, quarter, and year), the first set, the fifth set, and the sixth set can be determined first.

[0152] The first set includes each information item of the current transaction, that is, the first set includes the payer account number, the payee account number, the terminal information of the payment terminal and the IP information of the payment terminal of the current transaction.

[0153] The fifth set includes multiple preset second time periods, including weeks, months, quarters and years.

[0154] Among them, the sixth set includes multiple second transaction statistical dimensions, which are the total number of related transactions, the total transaction amount of related transactions, the total number of related transactions with successful payments, and the total amount of related transactions with successful payments.

[0155] Further, the Cartesian product of the first set, the fifth set, and the sixth set may be determined to obtain a plurality of third characteristic dimensions, each of which is used to characterize a fourth transaction statistical dimension of an information item of the current transaction within a second preset time period. For example, the third characteristic dimension may be used to characterize the total number of associated transactions of the payer account of the current transaction within a month, the total number of associated transactions of the payer account with successful payments within a quarter, and the like.

[0156] Based on this, the statistical data corresponding to each third characteristic dimension can be determined from the transaction database, and then the total number of related transactions, the total transaction amount of related transactions, the total number of related transactions with successful payments, and the total amount of related transactions with successful payments for each information item associated with the current transaction in each second preset time period can be obtained based on the statistical data of each third characteristic dimension.

[0157] In the embodiment of the present application, the payment environment change feature of the current transaction is used to characterize the change of the payment environment of the current transaction. The payment environment change feature of the current transaction may include at least one of the following:

[0158] first information used to indicate whether the terminal information of the payment terminal changes during the process of requesting to obtain the payment link;

[0159] Second information used to indicate whether the IP information of the payment terminal changes during the process of requesting to obtain the payment connection;

[0160] Third information used to indicate whether the browser user agent changes during the process of requesting to obtain the payment link;

[0161] The duration from the time the current transaction is initiated to the time the payment information is reported.

[0162] Among them, the transaction initiation time and the payment information reporting time can be determined based on the timestamps of each link in the payment information reported by the transaction initiator.

[0163] It should be noted that the payment environment change characteristics of the current transaction are only examples. For example, it may also include fourth information on whether the payment network of the current transaction has changed and fifth information on whether the payment method of the current transaction has changed, etc. The specific information can be determined based on the actual application scenario requirements and is not limited here.

[0164] In the embodiment of the present application, each information item combination includes at least two of all information items associated with the current transaction. For example, when all information items associated with the current transaction include the payer account, the payee account, the terminal information of the payment terminal, and the IP information of the payment terminal corresponding to the current transaction, the information item combination can be any two, three, or four of the payer account, the payee account, the terminal information of the payment terminal, and the IP information of the payment terminal, which can be determined based on the actual application scenario requirements and is not limited here.

[0165] Furthermore, the third transaction statistics dimension may include at least one of the following:

[0166] The total number of related transactions for which transaction subject verification is performed;

[0167] The total number of related transactions verified by the transaction subject;

[0168] The frequency of transactions per preset period.

[0169] The above-mentioned preset period may include annually, monthly or daily, etc., which is not limited here.

[0170] It should be noted that the third transaction statistical dimension mentioned above is only an example. For example, the third transaction statistical dimension can also be the total number of related transactions that have not been verified by the transaction subject, etc. The specific number can be determined based on the actual application scenario requirements and is not limited here.

[0171] For each information item combination of the current transaction, the associated transaction corresponding to each information item combination is a transaction associated with each information item of the information item combination. For example, if an information item combination of the current transaction consists of a payee account and a payee account of the current transaction, then the transaction associated with both the payee account and the payee account is the associated transaction corresponding to the information item combination.

[0172] See also Figure 6 , Figure 6 FIG. 1 is another schematic diagram of a scenario for determining transaction abnormality assessment features provided by an embodiment of the present application. Figure 6 As shown, when determining the statistical data corresponding to each third transaction statistical dimension for each information item and information item combination associated with the current transaction, the seventh set and the eighth set may be determined first.

[0173] The seventh set includes each information item of the current transaction, that is, the first set includes the payer account, the payee account, the terminal information of the payment terminal and the IP information of the payment terminal of the current transaction. And the seventh set also includes at least one information item combination composed of different information items.

[0174] Among them, the eighth set includes each third transaction statistical dimension, which are the total number of related transactions verified by the transaction subject, the total number of related transactions verified by the transaction subject, and the transaction frequency of each preset period.

[0175] Further, the Cartesian product of the seventh set and the eighth set may be determined to obtain multiple fourth characteristic dimensions, each of which is used to characterize an information item of the current transaction or a third transaction statistical dimension of a combination of information items. For example, the fourth characteristic dimension may be used to characterize the total number of associated transactions for which the payer account of the current transaction is verified by the transaction subject, and may also be used to characterize the total number of associated transactions for which the payer account and the payee account of the current transaction are verified by the transaction subject, and so on.

[0176] Based on this, the statistical data corresponding to each fourth characteristic dimension can be determined from the transaction database, and then the statistical data corresponding to each third transaction statistical dimension of each information item and information item combination associated with the current transaction can be obtained based on the statistical data of each fourth characteristic dimension.

[0177] In the embodiment of the present application, the transaction abnormality indicator corresponding to the payer of the current transaction may include at least one of the following:

[0178] The total number of IP information of the payer;

[0179] The total number of pre-set risk IP information in the payer's IP information;

[0180] The total number of recipients;

[0181] The total number of preset risk payees.

[0182] Among them, the total number of IP information of the payer is used to characterize the total number of IP information (that is, the IP information of the payment terminal used) used by the payer corresponding to the current transaction in all historical transactions. The preset risk IP information is the IP information of the payer in the historical transactions with malicious transaction behaviors, or the IP information of the payer in the historical transactions that have not been verified by the transaction subject, etc. The total number of preset risk IPs is used to characterize the total number of preset risk IPs involved in all historical transactions corresponding to the payer of the current transaction. The total number of payees is used to characterize the total number of payees involved in all historical transactions corresponding to the payer corresponding to the current transaction. The preset risk payee is the payee corresponding to the historical exchange with malicious transaction behaviors, or the payee corresponding to the historical exchange that has not been verified by the transaction subject, etc. The total number of preset risk payees is used to characterize the total number of preset risk payees involved in all historical transactions corresponding to the payee of the current transaction.

[0183] The transaction abnormality indicator corresponding to the payee of the current transaction may include at least one of the following:

[0184] The total number of IP information of the payer;

[0185] The total number of pre-set risk IP information in the payer's IP information;

[0186] The total number of payers;

[0187] The total number of pre-set risk payers.

[0188] Among them, the total number of IP information of payers is used to characterize the total number of IP information (that is, the IP information of the payment terminal used) used by payers in all historical transactions corresponding to the payee corresponding to the current transaction. The preset risk IP information is the IP information used by payers in historical transactions with malicious transaction behaviors, or the IP information corresponding to payers in historical transactions that have not been verified by the transaction subject, etc. The total number of preset risk IPs is used to characterize the total number of preset risk IPs in the IP information of payers in all historical transactions corresponding to the payer of the current transaction. The total number of payers is used to characterize the total number of payers in all historical transactions corresponding to the payee corresponding to the current transaction. The preset risk payer is the payer corresponding to the historical exchange with malicious transaction behaviors, or the payer corresponding to the historical exchange that has not been verified by the transaction subject, etc. The total number of preset risk payers is used to characterize the total number of preset risk payers involved in all historical transactions corresponding to the payee of the current transaction.

[0189] Step S23: determine the account anomaly coefficient of the current transaction based on the account anomaly assessment feature, determine the transaction anomaly coefficient of the current transaction based on the transaction anomaly assessment feature, and determine the transaction risk level of the current transaction based on the account anomaly coefficient and the transaction anomaly coefficient.

[0190] In some feasible implementations, after determining the account abnormality assessment feature of the current account based on the payment information, the account abnormality coefficient of the current transaction can be determined based on the account abnormality assessment feature. Determining the account abnormality coefficient of the current transaction based on the account abnormality assessment feature can be performed specifically through an account abnormality assessment model.

[0191] Among them, the account abnormality coefficient of the current transaction is used to characterize the risk level of the account associated with the current transaction, and the account risk coefficient is positively correlated with the risk level of the account associated with the current transaction.

[0192] Specifically, when determining the account anomaly coefficient of the current transaction based on the account anomaly assessment model, the account anomaly assessment feature can be represented by a multidimensional matrix to obtain a first initial feature, and the first initial feature is input into the account anomaly assessment model to obtain the account anomaly coefficient of the current transaction.

[0193] Among them, the account anomaly assessment model can extract and reduce the first initial feature based on the time series model to obtain a one-dimensional feature vector, and process the first feature vector through the classification model to obtain the account anomaly coefficient of the current transaction.

[0194] Among them, classification model refers to a type of algorithm in machine learning, such as the classification algorithm based on XGBoost, whose input data is fixed-dimensional features and output data is discrete labels. Time series model refers to a type of algorithm in machine learning, such as the algorithm based on LSTM network, which is used to extract features from data with time series features. The extracted features can be used for various downstream tasks such as classification and regression.

[0195] For example, see Figure 7 , Figure 7 Schematic diagram of a scenario for determining an abnormal coefficient of an account provided in an embodiment of the present application. Figure 7 As shown, when the account abnormality assessment features of the current transaction include the statistical data of each information item associated with the previous transaction corresponding to each first transaction statistical dimension in each time window, the account abnormality assessment features of the current account can be expressed as a 3-dimensional matrix of time_window_num*id_num*feat_num.

[0196] Among them, time_window_num represents the number of time windows, id_num represents the number of information items, and feat_num represents the number of the first transaction statistical dimension.

[0197] The account anomaly assessment model can perform feature extraction on the first initial model (the aforementioned 3D matrix) based on the time series model to obtain a feature matrix with a dimension of id_num*feat_num, and perform dimensionality reduction processing on the feature matrix to obtain a 1D feature vector. Finally, the account anomaly coefficient in the range of 0-1 is obtained through the classification model.

[0198] Optionally, when determining the account anomaly coefficient of the current transaction based on the account anomaly assessment feature of the current transaction, the account attention feature can be determined, and then the account anomaly assessment feature of the current transaction is represented by a multidimensional matrix to obtain a first initial feature. Thus, through the account anomaly assessment model, the first initial feature is subjected to attention fusion based on the account attention feature to obtain a first account fusion feature, and then the first account fusion feature is subjected to feature dimensionality reduction to obtain the final first feature vector, and the first feature vector is processed through the classification model to obtain the account anomaly coefficient of the current transaction.

[0199] When determining the account attention feature, the account abnormality assessment feature of the first target transaction may be represented by a multi-dimensional matrix to obtain a second initial feature, and the second initial feature may be used as the account attention feature.

[0200] The first target transaction is the last transaction initiated by the first transaction initiator, and the first transaction initiator is the transaction responder of the transaction initiated by the transaction initiator of the current transaction most recently at the current moment. Alternatively, the first target transaction is the last transaction initiated by the second transaction initiator, and the second transaction initiator is the transaction initiator corresponding to the last time the transaction initiator of the current transaction acted as the transaction responder.

[0201] Alternatively, when determining the account attention feature, a first data snapshot can be constructed based on at least one of the account abnormality assessment feature of the current transaction and the transaction abnormality assessment feature of the current transaction, and an image feature of the first data snapshot can be determined, and then the image feature can be determined as the account attention feature.

[0202] Alternatively, when determining the account attention feature, various features of the transaction anomaly assessment feature of the current transaction may be spliced ​​together to obtain a spliced ​​feature, and the spliced ​​feature may be determined as the account attention feature.

[0203] In some feasible implementations, after determining the transaction anomaly assessment feature of the current account based on the payment information, the transaction anomaly coefficient of the current transaction can be determined based on the transaction anomaly assessment feature. Determining the transaction anomaly coefficient of the current transaction based on the transaction anomaly assessment feature can be performed specifically through a transaction anomaly assessment model.

[0204] Among them, the transaction anomaly coefficient of the current transaction is used to characterize the risk level of the current transaction itself. The transaction risk coefficient is positively correlated with the risk level of the current transaction itself. The higher the risk level of the current transaction itself, the greater the possibility that the current transaction is initiated by the program subject.

[0205] Specifically, when determining the transaction anomaly coefficient of the current transaction based on the transaction anomaly assessment model, the various features within the transaction anomaly assessment feature can be spliced ​​to obtain a spliced ​​feature. After the spliced ​​feature is input into the transaction anomaly assessment model, the transaction anomaly assessment model performs dimensionality reduction on the spliced ​​feature, and the reduced dimensionality feature is processed by the classification model to obtain the transaction anomaly coefficient of the current transaction.

[0206] For example, see Figure 8 , Figure 8 Schematic diagram of the scenario of the abnormal coefficient of the transaction account provided in the embodiment of the present application. Figure 8 As shown, the transaction abnormality assessment features of the current transaction may include a first feature, a second feature, a third feature, and a fourth feature.

[0207] Among them, the first feature is the statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period; the second feature is the payment environment change feature of the current transaction; the third feature is the statistical data of each information item and information item combination associated with the current transaction corresponding to each third transaction statistical dimension; the fourth feature is the transaction abnormality indicator corresponding to at least one of the payer or payee of the current transaction.

[0208] The first feature, the second feature, the third feature and the fourth feature are concatenated and input into the transaction anomaly assessment model. The transaction anomaly assessment model reduces the dimension of the concatenated features to obtain a one-dimensional feature vector, and the one-dimensional feature vector is processed by the classification model to finally obtain the transaction anomaly coefficient (0-1) of the current transaction.

[0209] Optionally, when determining the account anomaly coefficient of the current transaction based on the transaction anomaly assessment feature of the current transaction, the transaction attention feature can be determined, and then the features in the transaction anomaly assessment feature are spliced ​​to obtain the spliced ​​feature. Thus, through the account anomaly assessment model, the spliced ​​feature is fused based on the account attention feature to obtain the first transaction fusion feature, and then the first transaction fusion feature is reduced in feature dimension, and the reduced dimension feature is processed through the classification model to obtain the transaction anomaly coefficient of the current transaction.

[0210] When determining the transaction attention feature, a first initial feature can be obtained by using a multi-dimensional matrix to represent the abnormality assessment feature of the account of the current transaction, and the first initial feature can be determined as the attention feature.

[0211] Alternatively, when determining the transaction attention feature, the account abnormality assessment feature of the first target transaction may be represented by a multi-dimensional matrix to obtain a second initial feature, and the second initial feature may be used as the transaction attention feature.

[0212] The first target transaction is the last transaction initiated by the first transaction initiator, and the first transaction initiator is the transaction responder of the transaction initiated by the transaction initiator of the current transaction most recently at the current moment. Alternatively, the first target transaction is the last transaction initiated by the second transaction initiator, and the second transaction initiator is the transaction initiator corresponding to the last time the transaction initiator of the current transaction acted as the transaction responder.

[0213] Alternatively, when determining the transaction attention feature, each feature in the transaction anomaly assessment feature of the second target transaction may be spliced ​​to obtain a spliced ​​feature, and the spliced ​​feature may be determined as the transaction attention feature.

[0214] The second target transaction is the transaction initiated last time by the transaction initiator of the current transaction.

[0215] Alternatively, when determining the transaction attention feature, a first data snapshot can be constructed based on at least one of the account abnormality assessment feature of the current transaction and the transaction abnormality assessment feature of the current transaction, and an image feature of the first data snapshot can be determined, and then the image feature can be determined as the transaction attention feature.

[0216] In some feasible implementations, after determining the account abnormality coefficient and transaction abnormality coefficient of the current transaction, the account abnormality risk weight and the transaction abnormality risk weight can be determined, thereby determining the weighted sum of the account abnormality coefficient and the transaction abnormality coefficient based on the account risk weight and the transaction risk weight, and determining the weighted sum as the transaction risk level of the current transaction.

[0217] Optionally, after determining the account anomaly coefficient and the transaction anomaly coefficient of the current transaction, the sum of the account anomaly coefficient and the transaction anomaly coefficient of the current transaction may be determined as the transaction risk level of the current transaction.

[0218] Optionally, after determining the account anomaly coefficient and the transaction anomaly coefficient of the current transaction, the sum or weighted sum of the account anomaly coefficient and the transaction anomaly coefficient of the current transaction is determined as the transaction risk coefficient, and based on the mapping relationship between the preset transaction risk coefficient and the transaction risk level, the transaction risk level of the current transaction is determined.

[0219] In some feasible implementations, after determining the transaction anomaly coefficient of the current transaction, the total number of transactions of the transaction initiator of the current transaction whose transaction risk level is greater than the preset risk threshold within the third preset time period may be determined. If the total number of transactions of the transaction initiator of the current transaction whose transaction risk level is greater than the preset risk threshold within the third preset time period exceeds the preset threshold, the first risk coefficient of the current transaction is multiplied by the historical risk weight to finally obtain the transaction risk coefficient of the current transaction.

[0220] Among them, the first risk coefficient of the current transaction is the sum or weighted sum of the account abnormality coefficient and the transaction abnormality coefficient of the current transaction.

[0221] At this time, the historical risk weight can be determined based on the mapping relationship between the risk weight and the total number of transactions with a risk level greater than a preset risk threshold, or the historical risk weight can also be a preset weight, which is not limited here.

[0222] Alternatively, if the proportion of transactions in which the transaction risk level of the transaction initiator of the current transaction is greater than the preset risk threshold within the third preset time period exceeds the proportion threshold of all transactions initiated by the transaction initiator of the current transaction within the third preset time period, the first risk coefficient of the current transaction is multiplied by the historical risk weight to finally obtain the transaction risk coefficient of the current transaction.

[0223] At this time, the historical risk weight can be determined based on the mapping relationship between the proportion of transactions with a transaction risk level greater than a preset risk threshold to all transactions and the risk weight, or the historical risk weight can also be a preset weight, which is not limited here.

[0224] The third preset time period may be a historical time period of a preset length before the current moment, such as one month or one year before the current moment, or may be all time periods before the current moment.

[0225] After the transaction risk coefficient of the current transaction is determined, the transaction risk level of the current transaction may be determined based on a preset mapping relationship between the transaction risk coefficient and the transaction risk level.

[0226] In some feasible implementations, the training methods of the account anomaly assessment model and the transaction anomaly assessment model can be found in Fig. 9 , Fig. 9 It is a flowchart of the model training method provided in the embodiment of the present application.

[0227] Fig. 9 The model training method shown may specifically include the following steps:

[0228] Step S91, determine a training sample set, where the training sample set includes account abnormality sample evaluation features and transaction abnormality sample evaluation features of multiple historical transactions.

[0229] In an embodiment of the present application, the account abnormal sample evaluation features and transaction abnormal sample evaluation features of multiple historical transactions that have passed the transaction subject verification or have not been verified by the transaction subject can be determined from the transaction database, as well as the account abnormal sample evaluation features and transaction abnormal sample evaluation features of multiple historical transactions that have not passed the transaction subject verification can be determined, and based on this, a training sample set for model training can be constructed.

[0230] The method for determining the abnormal sample evaluation characteristics of each historical transaction account and the abnormal sample evaluation characteristics of each transaction can be found in Figure 2 The method for determining the account abnormality assessment features and the transaction abnormality assessment features shown in step S22 will not be repeated here.

[0231] Step S92: input the account abnormality sample evaluation features of each historical transaction into the first initial model to obtain the account abnormality sample coefficient of the corresponding historical transaction, and input the transaction abnormality sample evaluation features of each historical transaction into the second initial model to obtain the transaction abnormality sample coefficient of the corresponding historical transaction.

[0232] The model structure of the first initial model and the specific implementation method of determining the abnormal sample coefficient of the account number can be as follows: Figure 7 As shown, the model structure of the second initial model and the specific implementation method of determining the transaction abnormal sample coefficient can be as follows: Figure 8 As shown, no further description is given here.

[0233] Step S93: determine the actual transaction risk level of each historical transaction, determine the transaction risk level of the corresponding historical transaction based on the account abnormal sample coefficient and the transaction abnormal sample coefficient of each historical transaction, and determine the total training loss based on the transaction risk level of each historical transaction and the actual transaction risk level.

[0234] In some feasible implementations, the implementation method of determining the predicted transaction risk level based on the account abnormal sample coefficient and the transaction abnormal sample coefficient of each historical transaction can be found in Figure 2 The implementation method of determining the transaction risk level of the current transaction shown in step S23 is not repeated here.

[0235] The actual transaction risk level of each historical transaction may be pre-set, thereby determining the total training loss based on the predicted transaction risk level and the actual transaction risk level of each historical transaction.

[0236] Optionally, when determining the actual transaction risk level of a historical transaction, for each historical transaction, if the historical transaction has not been verified by the transaction subject or has passed the transaction subject verification, the actual transaction risk level of the historical transaction is determined to be a first preset level (such as set to 0). If the historical transaction has not passed the transaction subject verification, the actual transaction risk level of the historical transaction is determined to be a second preset level (such as set to 1).

[0237] For each historical transaction, after determining the predicted transaction risk level of the historical transaction, if the predicted transaction risk level of the historical transaction is less than the preset risk level, the predicted transaction risk level of the historical transaction is determined as the first preset level (such as set to 0); if the predicted transaction risk level of the historical transaction is greater than or equal to the preset risk level, the predicted transaction risk level of the historical transaction is determined as the second preset level (such as set to 1).

[0238] The total training loss may be further determined based on the actual transaction risk level (the first preset level or the second preset level) and the predicted transaction risk level (the first preset level or the second preset level) of each historical transaction.

[0239] The total training loss can be determined based on the cross entropy loss function or other loss functions, which are not limited here.

[0240] The model training of the first initial model and the second initial model is supervised training. To ensure better robustness, the training sample set can be reconstructed each time the model is trained.

[0241] Step S94, iteratively train the first initial model and the second initial model based on the training sample set until the total training loss meets the training end condition, and stop the training, and determine the first initial model and the second initial model when the training is stopped as the account anomaly assessment model and the transaction anomaly assessment model, respectively.

[0242] In an embodiment of the present application, the total training loss can be determined each time the first initial model and the second initial model are trained based on the training sample set, and the training is stopped when the total training loss meets the training end condition. At this time, the first initial model when the training is stopped is determined as the final account anomaly assessment model, and the second initial model when the training is stopped is determined as the final transaction anomaly assessment model.

[0243] The training end condition may be that the total training loss reaches convergence, or that the total training loss is less than a preset loss threshold, etc. The specific condition may be determined based on the actual application scenario requirements and is not limited here.

[0244] The training method of the account anomaly assessment model and the transaction anomaly assessment model in the embodiment of the present application, the transaction anomaly coefficient and the determination method of the account anomaly coefficient can be implemented based on the machine learning technology in artificial intelligence. Artificial intelligence is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science, which attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines so that the machines have the functions of perception, reasoning and decision-making. Among them, machine learning is a special study of how computers simulate or implement human learning behavior to acquire new knowledge or skills, reorganize the existing knowledge structure to continuously improve its own performance, so as to realize the training process of the above-mentioned account anomaly assessment model and transaction anomaly assessment model and the determination process of transaction anomaly coefficient and account anomaly coefficient.

[0245] Among them, the abnormal sample evaluation features of each historical transaction account and the abnormal sample evaluation features of transactions in the above training sample set can be stored in a preset storage space. The preset storage space can be a remote dictionary service (Redis) system, a database (Database), a cloud storage (cloud storage) or a blockchain (Blockchain), which can be determined based on the actual application scenario requirements and is not limited here.

[0246] Among them, the database can be regarded as an electronic file cabinet in short - a place to store electronic files, which can be a relational database (SQL database) or a non-relational database (NoSQL database), which is not limited here. In this application, it can be used to store the abnormal sample evaluation features of each historical transaction account in the above training sample set and the abnormal sample evaluation features of the transaction. Blockchain is a new application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, which is a string of data blocks generated by cryptographic methods. In the embodiment of the present application, each data block in the blockchain can store the abnormal sample evaluation features of each historical transaction account in the above training sample set and the abnormal sample evaluation features of the transaction. Cloud storage is a new concept extended and developed from the concept of cloud computing. It refers to the use of cluster applications, grid technology, and distributed storage file systems. The functions of a large number of different types of storage devices (storage devices are also called storage nodes) in the network are collected and work together through application software or application interfaces to jointly store the abnormal sample evaluation features of each historical transaction account in the above training sample set and the abnormal sample evaluation features of the transaction.

[0247] Step S24: Process the current transaction based on the transaction risk level.

[0248] In some feasible implementations, after determining the transaction risk level of the current transaction, it can be determined whether the transaction risk level is less than the preset risk level. If the transaction risk level of the current transaction is less than the preset risk level, it means that the current transaction is initiated by a non-program subject, and at this time, the payment connection can be directly sent to the transaction initiator.

[0249] If the transaction risk level of the current transaction is greater than or equal to the preset risk level, it means that the current transaction may be initiated by the program subject. At this time, it is necessary to verify the transaction subject of the transaction initiator of the current transaction, and send a payment link to the transaction initiator after the transaction initiator passes the transaction subject verification.

[0250] Among them, the above-mentioned transaction subject verification includes but is not limited to mobile phone verification code recognition, image or character calibration recognition (such as requiring the transaction initiator to enter the prompted characters or click on the corresponding image), gesture recognition, voiceprint recognition, iris recognition, fingerprint recognition and face recognition, etc. The specific verification can be determined based on the actual application scenario requirements and is not limited here.

[0251] Optionally, if the transaction risk level of the current transaction is greater than or equal to the preset risk level, the target transaction subject verification method adopted by the current transaction can be determined based on the mapping relationship between the transaction risk level and the transaction subject verification method. Then, the transaction subject verification is performed on the transaction initiator based on the target transaction subject verification method, and a payment link is issued to the transaction initiator after the transaction initiator passes the transaction subject verification.

[0252] In some feasible implementations, the transaction subject of the current transaction may be verified based on threshold encryption technology.

[0253] Specifically, when verifying the transaction subject of the current transaction, a verification request can be sent to the transaction initiator. The verification request includes the first signature information of the verification information. The first signature information is obtained after the transaction responder of the current transaction (such as a shopping platform, backend server, etc.) signs the verification information based on the threshold public key.

[0254] After receiving the verification request, the transaction initiator of the current transaction can decrypt the first signature information based on the first threshold private key of the transaction initiator to obtain the first decrypted fragment, further perform hash calculation on the first decrypted fragment to obtain the first summary information, and then use the public key of the transaction responder of the current transaction to encrypt the first summary information to obtain the second signature information. After obtaining the second signature information, the transaction initiator of the current transaction can send response information, which includes the first decrypted fragment and the second signature information.

[0255] If the response information sent by the transaction initiator of the current transaction is not received, it is determined that the transaction initiator has not passed the transaction subject verification.

[0256] If a response message is received from the transaction initiator of the current transaction, the second signature message is decrypted based on the private key of the transaction responder to obtain the second summary message, and the first decrypted fragment is hashed to obtain the first summary message.

[0257] When the first summary information and the second summary information are inconsistent, it indicates that the information content of the response information has been tampered with. In this case, it is determined that the transaction initiator has not passed the transaction subject verification.

[0258] When the first summary information and the second summary information are consistent, it indicates that the information content of the response information has not been tampered with. At this time, the first signature information can be decrypted based on the second threshold private key of the transaction responder to obtain the second decrypted fragment.

[0259] If the above verification information is restored based on the first decrypted fragment and the second decrypted fragment, it means that the above first decrypted fragment is obtained by the transaction initiator using its own correct threshold private key to decrypt the first signature information, thereby determining that the transaction initiator has passed the transaction subject verification.

[0260] Among them, in the threshold signature encryption system, a threshold public key is first generated, and each object can use the threshold public key to encrypt information. Secondly, a threshold private key is generated for each specified object. The ownership of the private key is jointly controlled by the specified object. The specified object decrypts the encrypted information based on the threshold private key, and only by aggregating and restoring the decrypted fragments of no less than the number of fragments can the original unencrypted information be obtained.

[0261] In the embodiment of the present application, the number of shards is 2, that is, the unencrypted verification information can be determined based on the decrypted shards of the transaction initiator and the transaction responder.

[0262] Combine the following Fig.10 The transaction processing method provided in the embodiment of the present application is further explained. Fig.10 Schematic diagram of the process framework of the transaction processing method provided in the embodiment of the present application. Fig.10 As shown, there is a step of issuing a payment link in the complete payment link. For the current transaction, before issuing the payment link of the current transaction, the payment information reported by the transaction initiator in the process of obtaining the payment link of the current transaction can be obtained. The payment information includes information associated with the current transaction in each link before issuing the payment link connection, and the payment information can be obtained in the previous step (step n) of issuing the payment link, and can also be obtained in each step before issuing the payment link.

[0263] Furthermore, after obtaining the payment information, the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction can be determined based on the payment information, and the account abnormality assessment characteristics can be input into the account abnormality assessment model to obtain the account abnormality coefficient of the current transaction. At the same time, the transaction abnormality assessment characteristics can be input into the transaction abnormality assessment model to obtain the number of transaction abnormalities of the current transaction, and finally the transaction risk level of the current transaction can be determined based on the account abnormality coefficient and transaction abnormality coefficient of the current transaction.

[0264] Furthermore, if the transaction risk level of the current transaction is less than the preset risk threshold, a payment link is issued to the transaction initiator of the current transaction. If the transaction risk level of the current transaction is greater than or equal to the preset risk level, the transaction subject verification is performed on the transaction initiator of the current transaction, and a payment link is issued to the transaction initiator after the transaction initiator passes the transaction subject verification, otherwise the transaction subject verification is continuously performed on the transaction initiator until the transaction initiator passes the transaction subject verification or abandons the transaction subject verification.

[0265] The transaction processing method provided in the embodiment of the present application can determine the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction respectively through the payment information reported by the transaction initiator in the process of obtaining the payment link. Based on the account abnormality coefficient determined by the account abnormality assessment characteristics and the transaction abnormality coefficient determined by the transaction abnormality assessment characteristics, the transaction risk level of the current transaction can be accurately determined. Therefore, based on the transaction risk level, the transaction subject of the transaction that may be initiated by the program subject is verified, and the transaction subject of the transaction that may have transaction risks is verified on the payment link, effectively intercepting the behavior of the program subject to pull the payment link, improving transaction security, and eliminating potential transaction risks.

[0266] Furthermore, the transaction processing method provided in the embodiment of the present application can be embedded in the payment link of any transaction system or backend server, and is highly practical.

[0267] It should be noted that the payment information in the embodiments of this application is obtained in strict accordance with the requirements of relevant national laws and regulations, and the informed consent or separate consent of the transaction initiator is obtained, and the subsequent data use and processing behavior is carried out within the scope of authorization of laws and regulations and the transaction initiator. In addition, the face (or other biometric) recognition technology involved in the verification of the transaction subject in the embodiments of this application, when the above embodiments of this application are applied to specific products or technologies, the relevant data collection, use and processing process should comply with the requirements of national laws and regulations, and the information processing rules should be informed before collecting biometric information and the separate consent of the target object (transaction initiator) should be sought, and the biometric information should be processed in strict accordance with the requirements of laws and regulations and the personal information processing rules, and technical measures should be taken to ensure the security of relevant data.

[0268] See also Fig.11 , Fig.11 1 is a schematic diagram of the structure of a transaction processing device provided in an embodiment of the present application. The transaction processing device provided in an embodiment of the present application includes:

[0269] The information acquisition module 111 is used to acquire the payment information reported by the transaction initiator in the process of requesting to acquire the payment link of the current transaction;

[0270] A feature determination module 112, configured to determine an abnormal account assessment feature and an abnormal transaction assessment feature of the current transaction based on the payment information;

[0271] The level determination module 113 is used to determine the account abnormality coefficient of the current transaction based on the above-mentioned account abnormality assessment characteristics, determine the transaction abnormality coefficient of the current transaction based on the above-mentioned transaction abnormality assessment characteristics; determine the transaction risk level of the current transaction based on the above-mentioned account abnormality coefficient and the above-mentioned transaction abnormality coefficient;

[0272] The transaction processing module 114 is used to process the current transaction based on the above transaction risk level.

[0273] In some feasible implementations, the above-mentioned account abnormality assessment feature includes statistical data of each information item associated with the current transaction corresponding to each first transaction statistical dimension in each time window;

[0274] The information items associated with the current transaction include at least one of the following:

[0275] Payer account number;

[0276] Recipient's account number;

[0277] Terminal information of the payment terminal;

[0278] IP information of the payment terminal;

[0279] Among them, each of the above time windows is continuous and located within a first preset time period before the current transaction initiation moment.

[0280] In some feasible implementations, the first transaction statistical dimension includes at least one of the following:

[0281] The total number of related-party transactions;

[0282] The total transaction amount of related-party transactions;

[0283] The total number of related transactions for which transaction subject verification is performed;

[0284] The total number of related-party transactions for which verification of the transaction subject has not been carried out;

[0285] The total number of related transactions verified by the transaction subject;

[0286] The total number of related transactions that failed to pass the verification of the transaction entity.

[0287] In some feasible implementations, the transaction anomaly assessment feature includes at least one of the following:

[0288] Statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period;

[0289] The payment environment change characteristics of the current transaction, which are used to characterize the change status of the payment environment of the current transaction;

[0290] Each information item and information item combination associated with the current transaction corresponds to statistical data of each third transaction statistical dimension, and each of the above information item combinations includes any two of the above information items;

[0291] The transaction abnormality indicator corresponding to at least one of the payer or payee of the current transaction.

[0292] In some feasible implementations, the second transaction statistics dimension includes at least one of the following:

[0293] The total number of related-party transactions;

[0294] The total transaction amount of related-party transactions;

[0295] The total number of associated transactions with successful payments;

[0296] The total transaction amount of the associated transactions that have been successfully paid;

[0297] The transaction time of the most recent related transaction from the time the current transaction was initiated;

[0298] The transaction amount of the most recent related transaction before the current transaction is initiated;

[0299] The third transaction statistics dimension includes at least one of the following:

[0300] The total number of related transactions for which transaction subject verification is performed;

[0301] The total number of related transactions verified by the transaction subject;

[0302] Trading frequency.

[0303] In some feasible implementations, the payment environment change feature includes at least one of the following:

[0304] first information used to indicate whether the terminal information of the payment terminal changes during the process of requesting to obtain the payment link;

[0305] Second information used to indicate whether the IP information of the payment terminal changes during the process of requesting to obtain the payment link;

[0306] Third information used to indicate whether the browser user agent changes during the process of requesting to obtain the payment link;

[0307] The duration from the time the current transaction is initiated to the time the above payment information is reported.

[0308] In some feasible implementations, the transaction abnormality indicator corresponding to the payer includes at least one of the following:

[0309] The total number of IP information of the payer;

[0310] The total number of pre-set risk IP information in the payer's IP information;

[0311] The total number of recipients;

[0312] The total number of pre-set risk payees;

[0313] The transaction abnormality indicator corresponding to the payee includes at least one of the following:

[0314] The total number of IP information of the payer;

[0315] The total number of pre-set risk IP information in the payer's IP information;

[0316] The total number of payers;

[0317] The total number of pre-set risk payers.

[0318] In some feasible implementations, the level determination module 113 is used to:

[0319] Determine the abnormal account risk weight and abnormal transaction risk weight;

[0320] Based on the account risk weight, the transaction risk weight, the account anomaly coefficient and the transaction anomaly coefficient, the transaction risk level of the current transaction is determined.

[0321] In some feasible implementations, the transaction processing module 114 is used to:

[0322] In response to the transaction risk level being lower than a preset risk level, issuing a payment link to the transaction initiator;

[0323] In response to the transaction risk level being greater than or equal to the preset risk level, the transaction initiator is subject to transaction subject verification, and a payment link is issued to the transaction initiator after the transaction initiator passes the transaction subject verification.

[0324] In some feasible implementations, the account anomaly coefficient of the current transaction determined based on the account anomaly assessment feature is determined based on an account anomaly assessment model, and the transaction anomaly coefficient of the current transaction determined based on the transaction anomaly assessment feature is determined based on a transaction anomaly assessment model;

[0325] The above account anomaly assessment model and the above transaction anomaly assessment model are obtained by training the training module based on the following method:

[0326] Determine a training sample set, the training sample set including account abnormal sample evaluation features and transaction abnormal sample evaluation features of multiple historical transactions;

[0327] Input the abnormal account sample evaluation feature of each of the above historical transactions into the first initial model to obtain the abnormal account sample coefficient of the corresponding historical transaction, and input the abnormal transaction sample evaluation feature of each of the above historical transactions into the second initial model to obtain the abnormal transaction sample coefficient of the corresponding historical transaction;

[0328] Determine the actual transaction risk level of each of the above historical transactions, and determine the predicted transaction risk level of the corresponding historical transaction based on the account abnormal sample coefficient and the transaction abnormal sample coefficient of each of the above historical transactions; determine the total training loss based on the predicted transaction risk level and the actual transaction risk level of each of the above historical transactions;

[0329] The first initial model and the second initial model are iteratively trained based on the training sample set until the total training loss meets the training end condition, and the first initial model and the second initial model at the time of stopping the training are respectively determined as the account anomaly assessment model and the transaction anomaly assessment model.

[0330] In a specific implementation, the transaction processing device can execute the above-mentioned Figure 2 and / or Fig. 9 For the implementation methods provided in each step, please refer to the implementation methods provided in the above steps for details, which will not be repeated here.

[0331] See also Fig.12 , Fig.12 Schematic diagram of the structure of the electronic device provided in the embodiment of the present application. Fig.12As shown, the electronic device 1200 in this embodiment may include: a processor 1201, a network interface 1204 and a memory 1205. In addition, the above-mentioned electronic device 1200 may also include: an object interface 1203, and at least one communication bus 1202. Among them, the communication bus 1202 is used to realize the connection and communication between these components. Among them, the object interface 1203 may include a display screen (Display), a keyboard (Keyboard), and the object interface 1203 may optionally include a standard wired interface and a wireless interface. The network interface 1204 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1205 may be a high-speed RAM memory, or it may be a non-volatile memory (NVM), such as at least one disk storage. The memory 1205 may optionally also be at least one storage device located away from the aforementioned processor 1201. As Fig.12 As shown, the memory 1205 as a computer-readable storage medium may include an operating system, a network communication module, an object interface module, and a device control application program.

[0332] exist Fig.12 In the electronic device 1200 shown, the network interface 1204 can provide a network communication function; the object interface 1203 is mainly used to provide an input interface for the object; and the processor 1201 can be used to call the device control application stored in the memory 1205 to achieve:

[0333] Obtain the payment information reported by the transaction initiator when requesting the payment link for the current transaction;

[0334] Determine the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction based on the above payment information;

[0335] Determine the account abnormality coefficient of the current transaction based on the above account abnormality assessment characteristics, determine the transaction abnormality coefficient of the current transaction based on the above transaction abnormality assessment characteristics; determine the transaction risk level of the current transaction based on the above account abnormality coefficient and the above transaction abnormality coefficient;

[0336] The current transaction is processed based on the above transaction risk level.

[0337] In some feasible implementations, the above-mentioned account abnormality assessment feature includes statistical data of each information item associated with the current transaction corresponding to each first transaction statistical dimension in each time window;

[0338] The information items associated with the current transaction include at least one of the following:

[0339] Payer account number;

[0340] Recipient's account number;

[0341] Terminal information of the payment terminal;

[0342] IP information of the payment terminal;

[0343] Among them, each of the above time windows is continuous and located within a first preset time period before the current transaction initiation moment.

[0344] In some feasible implementations, the first transaction statistical dimension includes at least one of the following:

[0345] The total number of related-party transactions;

[0346] The total transaction amount of related-party transactions;

[0347] The total number of related transactions for which transaction subject verification is performed;

[0348] The total number of related-party transactions for which verification of the transaction subject has not been carried out;

[0349] The total number of related transactions verified by the transaction subject;

[0350] The total number of related transactions that failed to pass the verification of the transaction entity.

[0351] In some feasible implementations, the transaction anomaly assessment feature includes at least one of the following:

[0352] Statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period;

[0353] The payment environment change characteristics of the current transaction, which are used to characterize the change status of the payment environment of the current transaction;

[0354] Each information item and information item combination associated with the current transaction corresponds to statistical data of each third transaction statistical dimension, and each of the above information item combinations includes any two of the above information items;

[0355] The transaction abnormality indicator corresponding to at least one of the payer or payee of the current transaction.

[0356] In some feasible implementations, the second transaction statistics dimension includes at least one of the following:

[0357] The total number of related-party transactions;

[0358] The total transaction amount of related-party transactions;

[0359] The total number of associated transactions with successful payments;

[0360] The total transaction amount of the associated transactions that have been successfully paid;

[0361] The transaction time of the most recent related transaction from the time the current transaction was initiated;

[0362] The transaction amount of the most recent related transaction before the current transaction is initiated;

[0363] The third transaction statistics dimension includes at least one of the following:

[0364] The total number of related transactions for which transaction subject verification is performed;

[0365] The total number of related transactions verified by the transaction subject;

[0366] Trading frequency.

[0367] In some feasible implementations, the payment environment change feature includes at least one of the following:

[0368] first information used to indicate whether the terminal information of the payment terminal changes during the process of requesting to obtain the payment link;

[0369] Second information used to indicate whether the IP information of the payment terminal changes during the process of requesting to obtain the payment link;

[0370] Third information used to indicate whether the browser user agent changes during the process of requesting to obtain the payment link;

[0371] The duration from the time the current transaction is initiated to the time the above payment information is reported.

[0372] In some feasible implementations, the transaction abnormality indicator corresponding to the payer includes at least one of the following:

[0373] The total number of IP information of the payer;

[0374] The total number of pre-set risk IP information in the payer's IP information;

[0375] The total number of recipients;

[0376] The total number of pre-set risk payees;

[0377] The transaction abnormality indicator corresponding to the payee includes at least one of the following:

[0378] The total number of IP information of the payer;

[0379] The total number of pre-set risk IP information in the payer's IP information;

[0380] The total number of payers;

[0381] The total number of pre-set risk payers.

[0382] In some feasible implementations, the processor 1201 is configured to:

[0383] Determine the abnormal account risk weight and abnormal transaction risk weight;

[0384] Based on the account risk weight, the transaction risk weight, the account anomaly coefficient and the transaction anomaly coefficient, the transaction risk level of the current transaction is determined.

[0385] In some feasible implementations, the processor 1201 is configured to:

[0386] In response to the transaction risk level being lower than a preset risk level, issuing a payment link to the transaction initiator;

[0387] In response to the transaction risk level being greater than or equal to the preset risk level, the transaction initiator is subject to transaction subject verification, and a payment link is issued to the transaction initiator after the transaction initiator passes the transaction subject verification.

[0388] In some feasible implementations, the account anomaly coefficient of the current transaction determined based on the account anomaly assessment feature is determined based on an account anomaly assessment model, and the transaction anomaly coefficient of the current transaction determined based on the transaction anomaly assessment feature is determined based on a transaction anomaly assessment model;

[0389] The account anomaly assessment model and the transaction anomaly assessment model are obtained by training the processor 1201 based on the following method:

[0390] Determine a training sample set, the training sample set including account abnormal sample evaluation features and transaction abnormal sample evaluation features of multiple historical transactions;

[0391] Input the abnormal account sample evaluation feature of each of the above historical transactions into the first initial model to obtain the abnormal account sample coefficient of the corresponding historical transaction, and input the abnormal transaction sample evaluation feature of each of the above historical transactions into the second initial model to obtain the abnormal transaction sample coefficient of the corresponding historical transaction;

[0392] Determine the actual transaction risk level of each of the above historical transactions, and determine the predicted transaction risk level of the corresponding historical transaction based on the account abnormal sample coefficient and the transaction abnormal sample coefficient of each of the above historical transactions; determine the total training loss based on the predicted transaction risk level and the actual transaction risk level of each of the above historical transactions;

[0393] The first initial model and the second initial model are iteratively trained based on the training sample set until the total training loss meets the training end condition, and the first initial model and the second initial model at the time of stopping the training are respectively determined as the account anomaly assessment model and the transaction anomaly assessment model.

[0394] It should be understood that in some feasible implementations, the processor 1201 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the type of device.

[0395] In a specific implementation, the electronic device 1200 can execute the above-mentioned functions through its built-in functional modules. Figure 2 and / or Fig. 9 For the implementation methods provided in each step, please refer to the implementation methods provided in the above steps for details, which will not be repeated here.

[0396] The present application also provides a computer-readable storage medium, which stores a computer program and is executed by a processor to implement Figure 2 and / or Fig. 9 For the methods provided in each step, please refer to the implementation methods provided in the above steps for details, which will not be repeated here.

[0397] The computer-readable storage medium may be an internal storage unit of the transaction processing device or electronic device provided in any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart memory card (smart media card, SMC), a secure digital (secure digital, SD) card, a flash card (flash card), etc. equipped on the electronic device. The computer-readable storage medium may also include a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc. Further, the computer-readable storage medium may also include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0398] The present invention provides a computer program product, which includes a computer program. The computer program is executed by a processor. Figure 2 and / or Fig. 9 The methods provided in each step.

[0399] The terms "first", "second", etc. in the claims, specification and drawings of the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or electronic device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or electronic devices. Mentioning "embodiment" in this article means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. Displaying the phrase at various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments. The term "and / or" used in the present application specification and the attached claims refers to any combination of one or more of the items listed in association and all possible combinations, and includes these combinations.

[0400] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Professional and technical personnel may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0401] The above disclosure is only the preferred embodiment of the present application, which cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A transaction processing method, characterized in that: The method comprises: Obtain the payment information reported by the transaction initiator when requesting the payment link for the current transaction; Determine the account abnormality assessment characteristics and transaction abnormality assessment characteristics of the current transaction based on the payment information; Determine the account abnormality coefficient of the current transaction based on the account abnormality assessment feature, determine the transaction abnormality coefficient of the current transaction based on the transaction abnormality assessment feature; determine the transaction risk level of the current transaction based on the account abnormality coefficient and the transaction abnormality coefficient; The current transaction is processed based on the transaction risk level.

2. The method according to claim 1, characterized in that The account abnormality assessment feature includes statistical data of each information item associated with the current transaction corresponding to each first transaction statistical dimension in each time window; The information items associated with the current transaction include at least one of the following: Payer account number; Recipient's account number; Terminal information of the payment terminal; IP information of the payment terminal; Wherein, each of the time windows is continuous and located within a first preset time period before the current transaction initiation moment.

3. The method according to claim 2, characterized in that The first transaction statistics dimension includes at least one of the following: The total number of related-party transactions; The total transaction amount of related-party transactions; The total number of related transactions for which transaction subject verification is performed; The total number of related-party transactions for which verification of the transaction subject has not been carried out; The total number of related transactions verified by the transaction subject; The total number of related transactions that failed to pass the verification of the transaction entity.

4. The method according to claim 2, characterized in that: The transaction abnormality assessment feature includes at least one of the following: Statistical data of each information item associated with the current transaction corresponding to each second transaction statistical dimension in each second preset time period; The payment environment change feature of the current transaction, which is used to characterize the change status of the payment environment of the current transaction; Each information item and information item combination associated with the current transaction corresponds to statistical data of each third transaction statistical dimension, and each of the information item combinations includes any two of the information items; The transaction abnormality indicator corresponding to at least one of the payer or payee of the current transaction.

5. The method according to claim 4, characterized in that The second transaction statistics dimension includes at least one of the following: The total number of related-party transactions; The total transaction amount of related-party transactions; The total number of associated transactions with successful payments; The total transaction amount of the associated transactions that have been successfully paid; The transaction time of the most recent related transaction from the time the current transaction was initiated; The transaction amount of the most recent related transaction before the current transaction is initiated; The third transaction statistics dimension includes at least one of the following: The total number of related transactions for which transaction subject verification is performed; The total number of related transactions verified by the transaction subject; Trading frequency.

6. The method according to claim 5, characterized in that The payment environment change characteristics include at least one of the following: first information used to indicate whether the terminal information of the payment terminal changes during the process of requesting to obtain the payment link; Second information used to indicate whether the IP information of the payment terminal changes during the process of requesting to obtain the payment link; Third information used to indicate whether the browser user agent changes during the process of requesting to obtain the payment link; The duration from the time the current transaction is initiated to the time the payment information is reported.

7. The method according to claim 4, characterized in that The transaction abnormality indicator corresponding to the payer includes at least one of the following: The total number of IP information of the payer; The total number of pre-set risk IP information in the payer's IP information; The total number of recipients; The total number of pre-set risk payees; The transaction abnormality indicator corresponding to the payee includes at least one of the following: The total number of IP information of the payer; The total number of pre-set risk IP information in the payer's IP information; The total number of payers; The total number of pre-set risk payers.

8. The method according to claim 1, characterized in that The determining the transaction risk level of the current transaction based on the account abnormality coefficient and the transaction abnormality coefficient includes: Determine the abnormal account risk weight and abnormal transaction risk weight; Based on the account risk weight, the transaction risk weight, the account anomaly coefficient and the transaction anomaly coefficient, a transaction risk level of the current transaction is determined.

9. The method according to claim 1, characterized in that: The processing of the current transaction based on the transaction risk level includes: In response to the transaction risk level being lower than a preset risk level, issuing a payment link to the transaction initiator; In response to the transaction risk level being greater than or equal to the preset risk level, the transaction initiator is subject to transaction subject verification, and a payment link is issued to the transaction initiator after the transaction initiator passes the transaction subject verification.

10. The method according to claim 1, characterized in that The determining of the account anomaly coefficient of the current transaction based on the account anomaly assessment feature is determined based on an account anomaly assessment model, and the determining of the transaction anomaly coefficient of the current transaction based on the transaction anomaly assessment feature is determined based on a transaction anomaly assessment model; The account anomaly assessment model and the transaction anomaly assessment model are trained based on the following method: Determine a training sample set, wherein the training sample set includes account abnormal sample evaluation features and transaction abnormal sample evaluation features of multiple historical transactions; Inputting the abnormal account sample evaluation feature of each of the historical transactions into the first initial model to obtain the abnormal account sample coefficient of the corresponding historical transaction, and inputting the abnormal transaction sample evaluation feature of each of the historical transactions into the second initial model to obtain the abnormal transaction sample coefficient of the corresponding historical transaction; Determine the actual transaction risk level of each of the historical transactions, determine the predicted transaction risk level of the corresponding historical transaction based on the account abnormal sample coefficient and the transaction abnormal sample coefficient of each of the historical transactions; determine the total training loss based on the predicted transaction risk level and the actual transaction risk level of each of the historical transactions; The first initial model and the second initial model are iteratively trained based on the training sample set until the total training loss meets the training end condition, and the first initial model and the second initial model at the time of stopping the training are respectively determined as the account anomaly assessment model and the transaction anomaly assessment model.

11. A transaction processing device, characterized in that: The device comprises: The information acquisition module is used to obtain the payment information reported by the transaction initiator when requesting to obtain the payment link of the current transaction; A feature determination module, used to determine an abnormal account assessment feature and an abnormal transaction assessment feature of the current transaction based on the payment information; A level determination module, used to determine the account abnormality coefficient of the current transaction based on the account abnormality assessment feature, determine the transaction abnormality coefficient of the current transaction based on the transaction abnormality assessment feature; determine the transaction risk level of the current transaction based on the account abnormality coefficient and the transaction abnormality coefficient; The transaction processing module is used to process the current transaction based on the transaction risk level.

12. An electronic device, characterized in that: comprising a processor and a memory, wherein the processor and the memory are connected to each other; The memory is used to store computer programs; The processor is configured to execute the method according to any one of claims 1 to 10 when calling the computer program.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method according to any one of claims 1 to 10.

14. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 10 is implemented.

Images