CN115222409A

CN115222409A - Payment risk identification method and device

Info

Publication number: CN115222409A
Application number: CN202210767140.6A
Authority: CN
Inventors: 傅幸; 朱冠胤; 刘佳; 周璟; 吕乐; 杨阳; 王维强
Original assignee: ANT Financial Hang Zhou Network Technology Co Ltd
Current assignee: ANT Financial Hang Zhou Network Technology Co Ltd
Priority date: 2022-07-01
Filing date: 2022-07-01
Publication date: 2022-10-21

Abstract

One or more embodiments of the present specification provide a method and an apparatus for identifying payment risks, where a target payment behavior data set of a pending payment event of a target user is obtained; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively; acquiring a target payment portrait data subset corresponding to a target payment event node to which a payment event to be processed belongs in a user payment portrait data set of a target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by utilizing a deep neural network model based on a first historical payment behavior data set of a target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under a target payment event node; and determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to the payment characteristic dimensions.

Description

Payment risk identification method and device

Technical Field

The present document relates to the field of internet technologies, and in particular, to a method and an apparatus for identifying a payment risk.

Background

At present, with the coming of the internet era, the internet is widely applied to daily study, work and life of people. People's daily affairs can be handled and presented through the internet, and online and offline payment events are becoming more and more popular. However, in the payment process, there may be a problem of low payment security, in order to ensure the payment security of the user, before the user confirms the payment request, risk identification is performed on the payment event based on the payment behavior data by using a pre-trained risk identification model, and when it is determined that no risk payment exists in the payment event, the user confirms the payment request and triggers execution of the final payment completion operation.

Disclosure of Invention

It is an object of one or more embodiments of the present specification to provide a method of identifying payment risks. The identification method of the payment risk comprises the following steps:

acquiring a target payment behavior data set of a payment event to be processed of a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively. Acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by using a deep neural network model based on the first historical payment behavior data set of the target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node. Determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to the payment characteristic dimensions.

It is an object of one or more embodiments of the present specification to provide an identification device of payment risk. The device for identifying payment risk comprises:

the first data acquisition module is used for acquiring a target payment behavior data set of a payment event to be processed of a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively. The second data acquisition module is used for acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by using a deep neural network model based on a first historical payment behavior data set of the target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node. And the risk identification result determining module is used for determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to each payment characteristic dimension.

It is an object of one or more embodiments of the present specification to provide a risk of payment identification device, including: a processor; and a memory arranged to store computer executable instructions.

The computer-executable instructions, when executed, cause the processor to obtain a target payment behavior data set of a pending payment event for a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively. Acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by using a deep neural network model based on the first historical payment behavior data set of the target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node. Determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to the payment characteristic dimensions.

It is an object of one or more embodiments of the present specification to provide a storage medium for storing computer-executable instructions. The executable instructions, when executed by the processor, obtain a target payment behavior data set of a pending payment event for a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively. Acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by using a deep neural network model based on a first historical payment behavior data set of the target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node. Determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to the payment characteristic dimensions.

Drawings

In order to more clearly illustrate one or more embodiments or prior art solutions of the present specification, the drawings that are needed in the description of the embodiments or prior art will be briefly described below, it is obvious that the drawings in the following description are only some of the embodiments described in one or more of the specification, and that other drawings can be obtained by those skilled in the art without inventive exercise.

Fig. 1 is a schematic application scenario of a risk payment identification system according to one or more embodiments of the present disclosure;

FIG. 2 is a first flowchart of a method for identifying risk of payment provided in one or more embodiments of the present disclosure;

FIG. 3 is a second flowchart of a method for identifying risk of payment provided in one or more embodiments of the present disclosure;

FIG. 4 is a schematic diagram illustrating an implementation principle of a user payment portrait data set determination process in the identification method of payment risk provided by one or more embodiments of the present specification;

FIG. 5 is a third flowchart of a method for identifying risk of payment provided in one or more embodiments of the present disclosure;

fig. 6 is a schematic block diagram of an identification apparatus for risk of payment according to one or more embodiments of the present disclosure;

fig. 7 is a schematic structural diagram of an identification device for risk of payment provided in one or more embodiments of the present specification.

Detailed Description

In order to make the technical solutions in one or more embodiments of the present disclosure better understood, the technical solutions in one or more embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in one or more embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of one or more embodiments of the present disclosure, but not all embodiments. All other embodiments that can be derived by a person skilled in the art from the embodiments given in one or more of the present specification without inventive step shall fall within the scope of protection of this document.

It should be noted that one or more embodiments and features of the embodiments in the present description may be combined with each other without conflict. Reference will now be made in detail to one or more embodiments of the disclosure, examples of which are illustrated in the accompanying drawings.

One or more embodiments of the present specification provide a method and an apparatus for identifying a payment risk, in which if a risk identification model trained in advance is used to identify a risk of a payment event to be processed based on a target payment behavior data set before responding to a payment request of a target user, the payment request of the user is responded only when it is determined that no risk payment exists in the payment event to be processed, and a final payment completion operation is triggered to be executed, and since the risk identification model consumes a long time in a risk identification process, response efficiency of a service end to a payment request of a user terminal is inevitably reduced; considering that if the payment risk identification process is advanced, timeliness of payment risk control is improved, but the problem that the payment request response efficiency is low due to the fact that the prediction result of the payment behavior data is inaccurate due to the fact that the real payment behavior data is incomplete, and further the risk identification is still required to be carried out at the last moment by utilizing a pre-trained risk identification model based on the real payment behavior data is solved, therefore, a user payment portrait data set of a target user is determined by utilizing a trained deep neural network model in advance, and the user payment portrait data set comprises user payment portrait data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment portrait data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the historical payment behavior habit of the user can be quickly identified to determine whether the current payment event to be processed has the payment risk, risk identification is not needed to be carried out by utilizing a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be executed, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring the payment safety.

Fig. 1 is a schematic view of an application scenario of a risk of payment identification system provided in one or more embodiments of the present specification, as shown in fig. 1, the system includes: the system comprises a background server and a user terminal, wherein the user terminal can be a mobile terminal such as a smart phone and a tablet personal computer, the user terminal can also be a terminal device such as a personal computer, the background server can be a risk identification server for identifying risks of payment time of a user, the background server can be an independent server or a server cluster consisting of a plurality of servers; in addition, the risk identification server may be the same as or different from the payment server, and for a situation that the risk identification server and the payment server belong to different servers, the payment server needs to send a target payment behavior data set of a payment event to be processed to the risk identification server, and then the risk identification server returns a payment risk identification result obtained based on the target payment behavior data set to the payment server; the risk identification and the payment processing are executed by the same server, and the specific process of the payment risk identification is as follows:

the background server determines a first historical payment behavior data subset generated under each preset payment event node in a first historical payment behavior data set aiming at each preset payment event node in advance;

the background server side determines a user payment portrait data subset corresponding to each preset payment event node based on a first historical payment behavior data subset corresponding to each preset payment event node by using a pre-trained deep neural network model; determining a user payment portrait data set of a target user based on the user payment portrait data subsets corresponding to the preset payment event nodes; each user payment portrait data subset comprises second payment characteristic data under each payment characteristic dimension corresponding to a certain preset payment event node;

the user terminal sends a payment request to the background server after detecting the payment confirmation operation of the target user; wherein, the payment request carries event identification information of the payment event to be processed;

the background server side acquires a target payment behavior data set of the payment event to be processed after receiving the payment request sent by the user terminal; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively;

the background server side is used for acquiring a target payment portrait data subset corresponding to a target payment event node to which a payment event to be processed belongs from a user payment portrait data set of a target user; the target payment portrait data subset comprises second payment characteristic data corresponding to payment characteristic dimensions under the target payment event node;

the background server side determines a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to each payment characteristic dimension; the payment risk identification result may include any one of a payment risk degree score indicating that the payment risk exists for the payment event to be processed, a payment risk degree score indicating that the payment event to be processed exists the payment risk, and specifically, the payment risk identification result may be determined to indicate that the payment event to be processed does not exist the payment risk or indicate that the payment event to be processed exists the payment risk according to a size relationship between the payment risk degree score and a preset score threshold, where the payment risk identification result is the payment risk degree score;

the background server side responds to the payment request and triggers and executes payment processing operation if the payment risk identification result represents that the payment event to be processed has no payment risk; if the payment risk identification result represents that the payment risk exists in the payment event to be processed, intercepting the payment request and sending payment risk prompt information to a corresponding user terminal; specifically, the payment risk prompt message may be a risk prompt message displayed on the payment page, and considering that the payment risk may be caused by the loss of the user terminal of the target user, the payment risk prompt message may also be an email message sent to an email box reserved by the target user, an instant messaging message sent by an instant messaging account reserved by the target user, a short message sent by a phone number reserved by the target user, and the like.

In the application scene, a user payment portrait data set of a target user is determined by utilizing a trained deep neural network model in advance, wherein the user payment portrait data set comprises user payment portrait data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment portrait data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the historical payment behavior habit of the user can be quickly identified to determine whether the current payment event to be processed has the payment risk, risk identification is not needed to be carried out by utilizing a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be executed, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring the payment safety.

Fig. 2 is a first flowchart of a method for identifying a payment risk according to one or more embodiments of the present disclosure, where the method in fig. 2 can be executed by the backend server in fig. 1, as shown in fig. 2, and the method at least includes the following steps:

s202, acquiring a target payment behavior data set of a payment event to be processed of a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively, and the first payment characteristic data can be a first payment behavior characteristic value (namely the current payment behavior characteristic value) generated aiming at a payment event to be processed;

specifically, if the payment characteristic dimension is a payee, the corresponding first payment behavior characteristic value is a payee name or an account, if the payment characteristic dimension is payment time, the corresponding first payment behavior characteristic value is current time information, if the payment characteristic dimension is geographical location information, the corresponding first payment behavior characteristic value is location information of the user terminal of the target user, and if the payment characteristic dimension is payment amount, the corresponding first payment behavior characteristic value is the amount to be paid.

S204, acquiring a target payment portrait data subset corresponding to a target payment event node to which a payment event to be processed belongs from a user payment portrait data set of a target user; the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node; specifically, the user payment portrait data set is obtained by recalling payment characteristic data in advance by utilizing a deep neural network model based on a first historical payment behavior data set of a target user;

specifically, the second payment characteristic data may be a payment behavior characteristic vector including at least one second payment behavior characteristic value; each second payment behavior characteristic value refers to an optional payment behavior characteristic value under a certain payment characteristic dimension; for example, if the payment characteristic dimension is payee, the corresponding second payment behavior characteristic value may include payee 1, payee 2, \ 8230, payee N1, if the payment characteristic dimension is payment time, the corresponding second payment behavior characteristic value may include payment period 1, payment period 2, \8230, payment period N2, if the payment characteristic dimension is the geographic position information, the corresponding second payment behavior characteristic value may include a geographic position 1, a geographic position 2, \ 8230, and a geographic position N3, and if the payment characteristic dimension is the payment amount, the corresponding second payment behavior characteristic value may include a payment amount interval 1, a payment amount interval 2, \ 8230, and a payment amount interval N4.

The target payment event node may be one of a plurality of preset payment event nodes corresponding to a target user, for example, breakfast, lunch, and dinner may all be different preset payment event nodes; correspondingly, if the event occurrence of the payment event to be processed is located in the afternoon tea time, the target payment event node is a preset payment event node representing the afternoon tea, and if the event occurrence of the payment event to be processed is located in the lunch time, the target payment event node is a preset payment event node representing the lunch.

Specifically, the preset payment event node may be a conventional payment event node of a target user determined manually or automatically in advance, the preset payment event nodes corresponding to different users may be the same or different, and in order to improve accuracy of a user payment portrait data set, in specific implementation, a plurality of preset payment event nodes corresponding to the target user may be determined based on historical payment behavior data of the target user; then, for each preset payment event node, a deep neural network model is utilized to recall payment characteristic data based on a first historical payment behavior data set of a target user to obtain a corresponding user payment portrait data subset; each user payment image data subset can represent the conventional payment behavior habit of the target user under the preset payment event node, so that if the target payment behavior data set of the payment event to be processed is matched with the user payment image data subset under the target payment event node, the current payment event to be processed is in accordance with the historical payment behavior habit of the user.

S206, determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to the payment characteristic dimensions.

In specific implementation, the payment characteristic dimensions corresponding to different preset payment event nodes may be the same or different, for example, when comparing the current payment behavior characteristic value with a conventional payment behavior characteristic value (i.e., a selectable payment behavior characteristic value) for different preset payment event nodes, the considered payment characteristic dimensions may all be the same, partially different, or all be different; specifically, aiming at each payment characteristic dimension corresponding to a target payment event node, comparing first payment characteristic data and second payment characteristic data corresponding to the payment characteristic dimension to generate a payment behavior characteristic comparison result corresponding to the payment characteristic dimension; then, judging whether the payment event has payment risks or not by combining the payment behavior characteristic comparison results corresponding to the payment characteristic dimensions, and obtaining a payment risk identification result of the payment event to be processed; wherein the payment risk identification result may include: the payment risk exists in the payment event to be processed, or the payment risk does not exist in the payment event to be processed; the payment risk identification result may further include: and if the payment risk degree score is not greater than the preset score threshold, determining that the payment event to be processed has no risk payment.

In one or more embodiments of the present specification, a user payment representation data set of a target user is determined by using a trained deep neural network model in advance, where the user payment representation data set includes user payment representation data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment image data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the user historical payment behavior habit can be quickly identified to determine whether the payment event to be processed has a payment risk, risk identification is not needed to be carried out by using a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be carried out, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring payment safety.

In the process of determining the payment risk identification result, if the first payment feature data includes a first payment behavior feature value, and the second payment feature data is a payment behavior feature vector including a plurality of second payment behavior feature values, the step S206, based on the first payment feature data and the second payment feature data corresponding to each payment feature dimension, determines the payment risk identification result of the payment event to be processed, specifically includes:

step one, aiming at each payment characteristic dimension, determining a payment behavior characteristic comparison result corresponding to the payment characteristic dimension based on a first payment behavior characteristic value corresponding to the payment characteristic dimension and a plurality of second payment behavior characteristic values in a corresponding payment behavior characteristic vector;

for example, if the payment characteristic dimension is a payee, the payee X, which is indicated by the first payment behavior characteristic value and is aimed at the payment event to be processed, is compared with the payee 1, the payee 2, the payee 8230and the payee N1, which are indicated by the second payment behavior characteristic value corresponding to the target payment event node, to determine whether the payee X exists in the payee N1, if so, the payment characteristic dimension is determined to be that the comparison result of the payment behavior characteristics corresponding to the payee is characteristic data matching, and if so, the characteristic data is determined not matching; for another example, if the payment characteristic dimension is geographic position information, the occurring geographic position Y of the payment event to be processed represented by the first payment behavior characteristic value is compared with the geographic position 1, the geographic position 2, \8230andthe geographic position N3 represented by the second payment behavior characteristic value corresponding to the target payment event node, whether the geographic position 1, the geographic position 2, \8230existsin the geographic position N3 or not is determined, if yes, the payment characteristic dimension is determined that the payment behavior characteristic comparison result corresponding to the geographic position information is matched with the characteristic data, and if yes, the characteristic data is determined to be unmatched.

Determining a payment risk identification result of the payment event to be processed based on the payment behavior characteristic comparison result corresponding to each payment characteristic dimension;

specifically, the payment behavior feature comparison result corresponding to a certain payment feature dimension may be feature data matching or feature data mismatching, for example, if the payment behavior feature comparison results corresponding to all payment feature dimensions are feature data matching, or the number of the payment behavior feature comparison results matching the feature data is greater than a preset number threshold, it is determined that the payment risk identification result is that no risk payment exists for the payment event to be processed; for another example, according to the first quantity of the comparison result of the payment behavior characteristics matched with the characteristic data and the total number of the payment characteristic dimensions, a payment risk degree score representing the risk payment existing in the payment event to be processed is determined, and then the payment risk identification result of the payment event to be processed is determined according to the payment risk degree score.

Further, in the process of identifying payment risks for a payment event to be processed, a target payment representation data subset corresponding to a target payment event node needs to be used, so that a user payment representation data set of a target user needs to be determined in advance by using a trained deep neural network model, specifically, as shown in fig. 3, before acquiring a target payment behavior data set of a payment event to be processed of the target user in S202, the method further includes:

s208, acquiring a first historical payment behavior data set of the target user in a first historical time period;

specifically, the duration of the first historical time period may be set according to actual requirements, for example, the duration of the first historical time period may be one month, two months or half a year, and for a user with a low payment frequency, the duration of the first historical time period may be set to be greater, so that sufficient historical payment behavior data included in the first historical payment behavior data set can be ensured, and thus the accuracy of the user payment image data set is ensured; the first set of historical payment behavior data may include historical payment behavior data for a plurality of historical payment events, each historical payment event corresponding to a preset payment event node.

S210, inputting the first historical payment behavior data set into a pre-trained deep neural network model, and recalling payment characteristic data to obtain a user payment portrait data subset corresponding to each preset payment event node; the user payment portrait data subset comprises second payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively;

specifically, the first historical payment behavior data set may include historical payment behavior data of a plurality of historical payment events, each historical payment event corresponds to one preset payment event node, then, for each preset payment event node, the set of historical payment behavior data of the historical payment event belonging to the preset payment event node is determined as a first historical payment behavior data subset corresponding to the preset payment event node, and then, the user payment portrait data subset corresponding to the preset payment event node is determined by using the deep neural network model and based on the first historical payment behavior data subset.

It should be noted that, the deep neural network model may use the existing graph neural network model GNN, transformer sequence model, and the like, which can recall the first preset amount of target data (i.e., target data of topN in the plurality of candidate data) from the plurality of candidate data, and therefore, details are not repeated herein.

S212, generating a user payment portrait data set of a target user based on the user payment portrait data subsets corresponding to the preset payment event nodes; specifically, a set of user payment portrait data subsets corresponding to each preset payment event node can be directly determined as a user payment portrait data set of a target user; or, the user payment portrait data subsets corresponding to the preset payment event nodes may be subjected to preset processing (such as deduplication processing and optimization processing), and a set of the preset processed user payment portrait data subsets corresponding to the preset payment event nodes is determined as a user payment portrait data set of the target user.

In a process of determining the user payment profile data subset corresponding to each preset payment event node, in step S210, the first historical payment behavior data set is input to a pre-trained deep neural network model, and a payment feature data recall is performed to obtain the user payment profile data subset corresponding to each preset payment event node, which specifically includes:

step A1, aiming at each preset payment event node, determining a first historical payment behavior data subset belonging to the preset payment event node in a first historical payment behavior data set;

specifically, a plurality of historical payment events are determined based on a first historical payment behavior data set; determining a preset payment event node corresponding to each historical payment event based on a plurality of preset payment event nodes; for each preset payment event node, determining historical payment behavior data of a historical payment event belonging to the preset payment event node in a first historical payment behavior data set; and determining a first historical payment behavior data subset belonging to the preset payment event node based on the set of the historical payment behavior data corresponding to the preset payment event node.

Step A2, inputting a first historical payment behavior data subset corresponding to a preset payment event node into a pre-trained deep neural network model, and recalling payment characteristic data to obtain second payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively;

specifically, for each payment characteristic dimension corresponding to a certain preset payment event node, selecting a first preset number of second payment behavior characteristic values in the top sequence from a plurality of historical payment behavior characteristic values related to the payment characteristic dimension by using a pre-trained deep neural network model based on a first historical payment behavior data subset corresponding to the preset payment event node; and determining the payment behavior feature vector containing the first preset number of second payment behavior feature values as second payment feature data corresponding to the payment feature dimension.

And A3, generating a user payment portrait data subset corresponding to the preset payment event node based on second payment characteristic data corresponding to a plurality of payment characteristic dimensions corresponding to the preset payment event node respectively.

In the above step A2, the first historical payment behavior data subset corresponding to the preset payment event node is input to the pre-trained deep neural network model, and the payment feature data is recalled to obtain the second payment feature data corresponding to the plurality of payment feature dimensions, specifically including:

step A21, inputting a first historical payment behavior data subset corresponding to a preset payment event node into a pre-trained deep neural network model for each payment characteristic dimension, and recalling payment characteristic data to obtain a payment behavior characteristic vector corresponding to the payment characteristic dimension; the payment behavior feature vector comprises a first preset number of second payment behavior feature values which are ranked in the top;

specifically, in the process of determining the first preset number of second payment behavior characteristic values, the historical payment behavior characteristic values with the current number in the front sequence can be selected based on the occurrence times of the historical payment behavior characteristic values, so as to determine the first preset number of second payment behavior characteristic values (namely, all the historical payment behavior characteristic values); however, considering that the payment behavior of the target user may vary, taking the payment characteristic dimension as the payee as an example, although the payee B has a relatively small number of occurrences or a zero number of occurrences, if the payee B has a relatively high degree of association with the payee a (i.e., the payee belonging to the top several ranked digits), the payee B is also determined as the second payment behavior characteristic value of the first preset number ranked top, for example, if the payee B of the top several ranked digits includes the coffee shop a (i.e., the original payment behavior characteristic value), the coffee shop B near the coffee shop a and having similar consumption criteria (i.e., the associated payment behavior characteristic value) is also determined as the second payment behavior characteristic value of the first preset number ranked top, based on which, for each payment characteristic dimension, the second payment behavior characteristic value of the first preset number may include: the original payment behavior characteristic values of the second preset quantity and the associated payment behavior characteristic values of the third preset quantity; the sum of the second preset quantity and the third preset quantity is equal to the first preset quantity; and the correlation degree of the correlation payment behavior characteristic value and at least one original payment behavior characteristic value in the second preset number of original payment behavior characteristic values is greater than a preset threshold value.

Step a22, determining second payment characteristic data corresponding to the payment characteristic dimension based on the payment behavior characteristic vector corresponding to the payment characteristic dimension.

In a specific embodiment, the determination process of the user payment portrait data set about the target user, as shown in fig. 4, specifically includes:

acquiring a first historical payment behavior data set of a target user in a first historical time period;

determining a first historical payment behavior data subset corresponding to each preset payment event node based on the first historical payment behavior data set; for example, the payment behavior occurrence cycle is one day, and the preset payment event nodes include a preset payment event node 1 (e.g., early travel), a preset payment event node 2 (e.g., breakfast), a preset payment event node 3 (e.g., lunch), a preset payment event node 4 (e.g., afternoon tea), and a preset payment event node 5 (e.g., late shopping);

for each preset payment event node, inputting a first historical payment behavior data subset corresponding to each preset payment event node into a pre-trained deep neural network model;

for each payment characteristic dimension, sorting a plurality of historical payment behavior characteristic values corresponding to the payment characteristic dimension by using a deep neural network model and based on a first historical payment behavior data subset, and outputting corresponding second payment characteristic data, wherein the second payment characteristic data is a payment behavior characteristic vector containing a first preset number of second payment behavior characteristic values which are sorted in the payment characteristic dimension in an earlier way;

generating a user payment portrait data subset corresponding to the preset payment event node based on second payment characteristic data respectively corresponding to a plurality of payment characteristic dimensions corresponding to the preset payment event node;

and determining a user payment portrait data set of the target user based on the user payment portrait data subsets corresponding to the preset payment event nodes respectively.

Further, in order to improve the accuracy of the preset payment event node, thereby improving the accuracy of the target payment image data subset for payment risk identification, and further improving the accuracy of the result of payment risk identification, as shown in fig. 5, before acquiring the first historical payment behavior data set within the first historical time period of the target user in the above S208, the method further includes:

s214, acquiring a second historical payment behavior data set of the target user in a second historical time period;

the second historical time period may be the same as or different from the first historical time period, that is, the second historical payment behavior data set used for determining the preset payment event node may be the same as or different from the first historical payment behavior data set used for determining the user payment image data set of the target user; specifically, the duration of the second historical time period may also be set according to actual requirements, for example, the duration of the second historical time period may be one month, two months, or half a year, or even longer, and for a user with a low payment frequency, the duration of the second historical time period may also be set to be larger, so that sufficient historical payment behavior data included in the second historical payment behavior data set can be ensured, thereby ensuring the accuracy of a plurality of preset payment event nodes corresponding to the target user; the second set of historical payment behavior data may also include historical payment behavior data for a plurality of historical payment events.

S216, determining a plurality of preset payment event nodes corresponding to at least one payment behavior occurrence cycle type respectively based on the second historical payment behavior data set;

specifically, for each payment behavior occurrence cycle type i, determining, based on a second historical payment behavior data set, a plurality of historical payment events corresponding to Mi payment behavior occurrence cycles belonging to the payment behavior occurrence cycle type i in a second historical time period, where if the payment behavior occurrence cycle type of the target user is a "day type" (that is, the duration of each occurrence cycle in the occurrence cycle type is one day), the duration of the second historical time period is one month (taking 30 days in total as an example), M1=30, if the number of the payment behavior occurrence cycle types of the target user is 2 (that is, a working day and a holiday), the duration of the second historical time period is one month (4 weeks in total, 5 days of working day per week, 2 days of holiday as an example), M1=20 corresponding to the working day, and M2=4 corresponding to the holiday; the method comprises the steps of dividing historical payment events based on historical payment behavior data of a plurality of historical payment events corresponding to each payment behavior occurrence cycle type i, dividing the historical payment events with the similarity of the payment behavior data larger than a preset similarity threshold (such as the payment time interval is smaller than a preset time threshold, the geographic position distance is smaller than a preset distance threshold, and the like) into a historical event set, determining each historical event set as a preset payment event node, and in specific implementation, dividing the plurality of historical payment events into a plurality of cluster clusters based on the historical payment behavior data of the historical payment events by using a preset clustering method, and determining each cluster as a preset payment event node, namely, automatically determining a plurality of key payment event nodes which are conventionally executed by a target user in a payment behavior occurrence cycle under a certain payment behavior occurrence cycle type based on a second historical payment behavior data set.

Further, in order to improve the accuracy of the preset payment event node, the payment behavior occurrence cycle type may be automatically determined based on the historical payment behavior data, and then, for each payment behavior occurrence cycle type, the preset payment event node included in one payment behavior occurrence cycle of the payment behavior occurrence cycle type is determined, specifically, in step S214, based on the second historical payment behavior data set, a plurality of preset payment event nodes corresponding to at least one payment behavior occurrence cycle type are determined, and the method specifically includes:

step B21, determining at least one payment behavior occurrence cycle type based on the second historical payment behavior data set;

specifically, considering different users, the division results of the periodic payment events are different, and therefore, at least one payment behavior occurrence cycle type corresponding to the target user may be determined based on the second historical payment behavior data set of the target user; for example, the degree of similarity between the work-day payment behavior of the user a and the holiday payment behavior is higher, and therefore, the "day type" can be determined as one payment behavior occurrence cycle type of the user a, that is, in the second historical time period, the number of payment behavior occurrence cycles in the occurrence cycle type is the total number of days included in the second historical time period; for another example, the difference between the payment behavior of the user B on the working day and the payment behavior on the holiday is relatively large, so that the "working day cycle type" and the "holiday cycle type" may be respectively determined as different payment behavior occurrence cycle types of the user B, that is, in the second historical time period, the number of the payment behavior occurrence cycles of the occurrence cycle type on the working day is the number of days of the working day included in the second historical time period, and correspondingly, the number of the payment behavior occurrence cycles of the occurrence cycle type on the holiday is the number of days of the holiday included in the second historical time period; as another example, the payment events of the week of the user C have a certain periodicity, and therefore, the "week type" may be determined as a payment behavior occurrence cycle type of the user C, that is, in the second historical time period, the number of payment behavior occurrence cycles in the occurrence cycle type is the number of weeks included in the second historical time period.

Step B22, aiming at each payment behavior occurrence cycle type, determining a plurality of second historical payment behavior data subsets corresponding to the payment behavior occurrence cycle type in the second historical payment behavior data set; wherein a payment activity occurrence period under the payment activity occurrence period type may correspond to a second historical payment activity data subset;

step B23, based on a plurality of second historical payment behavior data subsets, determining a plurality of preset payment event nodes corresponding to the payment behavior occurrence cycle types, namely a plurality of conventional payment event nodes triggered and executed by the target user in a payment behavior occurrence cycle under a certain payment behavior occurrence cycle type;

specifically, in the process of determining a plurality of preset payment event nodes corresponding to each payment behavior occurrence cycle type, referring to the specific description of step B2, based on the second historical payment behavior data set, several key payment event nodes that are conventionally executed by the target user in one payment behavior occurrence cycle of a certain payment behavior occurrence cycle type are automatically determined.

Specifically, the preset payment event node may be a conventional payment event node of a target user determined manually or automatically in advance, and the preset payment event nodes corresponding to different users may be the same or different, and further, the payment behavior occurrence cycle type may also be a type to which a time cycle to which a plurality of conventional payment event nodes of the target user periodically occur belongs, which is determined manually or automatically in advance, where, for a case where the payment behavior occurrence cycle type is determined manually in advance, each day may be used as a payment behavior occurrence cycle type, each week may be used as a payment behavior occurrence cycle type, and each working day and each holiday may be used as different payment behavior occurrence cycle types.

In specific implementation, in order to improve the accuracy of the preset payment event node and thus improve the accuracy of the user payment portrait data set, at least one payment behavior occurrence cycle type corresponding to a target user can be determined based on historical payment behavior data of the target user; then, for each payment behavior occurrence cycle type, determining a plurality of preset payment event nodes corresponding to the payment behavior occurrence cycle type based on a plurality of second historical payment behavior data subsets corresponding to the payment behavior occurrence cycle type; next, in the process of identifying the payment risk, it may be determined which payment behavior occurrence cycle type the payment event to be processed belongs to, and then it may be determined which preset payment event node (i.e. target payment event node) under the payment behavior occurrence cycle type the payment event to be processed belongs to.

In a case where the target user corresponds to a plurality of payment behavior occurrence cycle types, in the user payment representation data set of the target user, in step S204, the obtaining of the target payment representation data subset corresponding to the target payment event node to which the payment event to be processed belongs specifically includes:

step one, determining a target payment behavior occurrence cycle type to which a payment event to be processed belongs based on event occurrence time of the payment event to be processed;

step two, determining a target payment event node to which a payment event to be processed belongs in a plurality of preset payment event nodes corresponding to the target payment behavior occurrence cycle type;

and step three, determining a target payment portrait data subset corresponding to the target payment event node in the user payment portrait data set of the target user.

Specifically, in the case that the number of the payment behavior occurrence periods corresponding to the target user is multiple, in the process of identifying the payment risk, which payment behavior occurrence period type (i.e., the target payment behavior occurrence period type) the payment event to be processed belongs to may be determined first, and then which preset payment event node (i.e., the target payment event node) the payment event to be processed belongs to under the payment behavior occurrence period type may be determined, so that the target payment portrait data subset corresponding to the target payment event node may be determined, and thus, the accuracy of the target payment portrait data subset may be improved.

In the method for identifying payment risks in one or more embodiments of the present specification, a user payment portrait data set of a target user is determined by using a trained deep neural network model in advance, where the user payment portrait data set includes user payment portrait data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment portrait data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the historical payment behavior habit of the user can be quickly identified to determine whether the current payment event to be processed has the payment risk, risk identification is not needed to be carried out by utilizing a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be executed, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring the payment safety.

Corresponding to the method for identifying a risk of payment described in fig. 2 to 5, based on the same technical concept, one or more embodiments of the present specification further provide an apparatus for identifying a risk of payment, fig. 6 is a schematic diagram of modules of the apparatus for identifying a risk of payment provided in one or more embodiments of the present specification, the apparatus is configured to perform the method for identifying a risk of payment described in fig. 2 to 5, as shown in fig. 6,

a first data obtaining module 602, configured to obtain a target payment behavior data set of a pending payment event of a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively;

a second data obtaining module 604, configured to obtain, in the user payment representation data set of the target user, a target payment representation data subset corresponding to a target payment event node to which the payment event to be processed belongs; the user payment sketch data set is obtained by recalling payment characteristic data in advance by using a deep neural network model based on a first historical payment behavior data set of the target user, and the target payment sketch data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node;

a risk identification result determining module 606, configured to determine a payment risk identification result of the to-be-processed payment event based on the first payment characteristic data and the second payment characteristic data corresponding to each of the payment characteristic dimensions.

In the device for identifying payment risks in one or more embodiments of the present specification, a user payment representation data set of a target user is determined by using a trained deep neural network model in advance, where the user payment representation data set includes user payment representation data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment portrait data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the historical payment behavior habit of the user can be quickly identified to determine whether the current payment event to be processed has the payment risk, risk identification is not needed to be carried out by utilizing a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be executed, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring the payment safety.

It should be noted that, the embodiment of the identification apparatus for risk of payment in this specification and the embodiment of the identification method for risk of payment in this specification are based on the same inventive concept, and therefore, specific implementation of this embodiment may refer to implementation of the aforementioned corresponding identification method for risk of payment, and repeated details are not repeated.

Further, on the basis of the same technical concept corresponding to the methods shown in fig. 2 to 5, one or more embodiments of the present specification further provide a risk payment identification device, which is configured to perform the above risk payment identification method, as shown in fig. 7.

The identification device of the payment risk may have a large difference due to different configurations or performances, and may include one or more processors 701 and a memory 702, where one or more stored applications or data may be stored in the memory 702. Memory 702 may be, among other things, transient storage or persistent storage. The application stored in memory 702 may include one or more modules (not shown), each of which may include a series of computer-executable instructions in an identification device of risk of payment. Still further, processor 701 may be configured to communicate with memory 702 to execute a series of computer-executable instructions in memory 702 on the risk of payment identification device. The identification means of payment risk may also comprise one or more power supplies 703, one or more wired or wireless network interfaces 704, one or more input output interfaces 705, one or more keyboards 706, etc.

In a particular embodiment, the identification of payment risks apparatus includes a memory, and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer-executable instructions in the identification of payment risks apparatus, and the one or more programs configured for execution by the one or more processors include computer-executable instructions for:

acquiring a target payment behavior data set of a payment event to be processed of a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively;

acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment sketch data set is obtained by recalling payment characteristic data in advance by using a deep neural network model based on a first historical payment behavior data set of the target user, and the target payment sketch data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node;

determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to each payment characteristic dimension.

In the identification device for payment risks in one or more embodiments of the present specification, a user payment representation data set of a target user is determined by using a trained deep neural network model in advance, where the user payment representation data set includes user payment representation data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment portrait data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the historical payment behavior habit of the user can be quickly identified to determine whether the current payment event to be processed has the payment risk, risk identification is not needed to be carried out by utilizing a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be executed, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring the payment safety.

It should be noted that, the embodiment of the identification device for a payment risk in this specification and the embodiment of the identification method for a payment risk in this specification are based on the same inventive concept, and therefore, for specific implementation of this embodiment, reference may be made to implementation of the corresponding identification method for a payment risk, and repeated parts are not described again.

Further, based on the same technical concept, one or more embodiments of the present specification further provide a storage medium for storing computer-executable instructions, where in a specific embodiment, the storage medium may be a usb disk, an optical disk, a hard disk, and the like, and when being executed by a processor, the storage medium stores the computer-executable instructions and can implement the following processes:

acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by utilizing a deep neural network model based on a first historical payment behavior data set of the target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node;

determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to the payment characteristic dimensions.

The storage medium in one or more embodiments of the present specification stores computer-executable instructions that, when executed by the processor, determine a user payment profile data set of a target user by pre-utilizing a trained deep neural network model, the user payment profile data set including user payment profile data subsets corresponding to a plurality of preset payment event nodes; then, in the process of payment risk identification, the first payment characteristic data of the current payment event to be processed is directly compared with the second payment characteristic data in the target payment image data subset corresponding to the target payment event node, so that whether the current payment event to be processed accords with the user historical payment behavior habit can be quickly identified to determine whether the payment event to be processed has a payment risk, risk identification is not needed to be carried out by using a risk identification model and based on the first payment characteristic data generated by the current payment event, namely when determining whether to respond to a payment request from a user terminal of a target user, only simple data comparison processing needs to be carried out, the identification efficiency of the payment risk is improved, and the response efficiency of a service end to the payment request of the user terminal is improved under the condition of ensuring payment safety.

It should be noted that the embodiment of the storage medium in this specification and the embodiment of the identification method of the payment risk in this specification are based on the same inventive concept, and therefore, for specific implementation of this embodiment, reference may be made to implementation of the aforementioned corresponding identification method of the payment risk, and repeated details are not described again.

The foregoing description of specific embodiments has been presented for purposes of illustration and description. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

In the 90's of the 20 th century, improvements to a technology could clearly distinguish between improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements to process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as ABEL (Advanced Boolean Expression Language), AHDL (alternate Hardware Description Language), traffic, CUPL (core universal Programming Language), HDCal, jhddl (Java Hardware Description Language), lava, lola, HDL, PALASM, rhyd (Hardware Description Language), and vhigh-Language (Hardware Description Language), which is currently used in most popular applications. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC625D, atmel AT91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the various elements may be implemented in the same one or more software and/or hardware implementations of one or more of the present descriptions.

As will be appreciated by one skilled in the art, one or more embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, one or more of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, one or more of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied in the medium.

One or more of the present specification has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to one or more embodiments of the specification. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of other like elements in a process, method, article, or apparatus comprising the element.

As will be appreciated by one skilled in the art, one or more embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, one or more of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, one or more of the present description can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

One or more of the present specification can be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. One or more of the present specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

The above description is merely illustrative of one or more embodiments of the present disclosure and is not intended to limit one or more embodiments of the present disclosure. Various modifications and alterations to one or more of the present descriptions will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of one or more of the present specification should be included in the scope of one or more claims of the present specification.

Claims

1. A method of identifying payment risks, comprising:

2. The method of claim 1, wherein the first payment characteristic data comprises a first payment behavior characteristic value and the second payment characteristic data is a payment behavior characteristic vector comprising a plurality of second payment behavior characteristic values;

determining a payment risk identification result of the to-be-processed payment event based on the first payment characteristic data and the second payment characteristic data corresponding to each payment characteristic dimension, including:

for each payment characteristic dimension, determining a payment behavior characteristic comparison result corresponding to the payment characteristic dimension based on the first payment behavior characteristic value corresponding to the payment characteristic dimension and a plurality of second payment behavior characteristic values in the corresponding payment behavior characteristic vector;

and determining a payment risk identification result of the payment event to be processed based on the payment behavior characteristic comparison result corresponding to each payment characteristic dimension.

3. The method of claim 1, wherein prior to obtaining the target payment behavior data set for the pending payment event for the target user, further comprising:

inputting the first historical payment behavior data set into a pre-trained deep neural network model, and recalling payment characteristic data to obtain a user payment portrait data subset corresponding to each preset payment event node; the user payment portrait data subset comprises second payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively;

and generating a user payment portrait data set of the target user based on the user payment portrait data subsets corresponding to the preset payment event nodes.

4. The method of claim 3, wherein the inputting the first historical payment behavior data set into a pre-trained deep neural network model for payment feature data recall to obtain a user payment profile data subset corresponding to each preset payment event node comprises:

for each preset payment event node, determining a first historical payment behavior data subset belonging to the preset payment event node in a first historical payment behavior data set;

inputting the first historical payment behavior data subset corresponding to the preset payment event node into a pre-trained deep neural network model, and recalling payment characteristic data to obtain second payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively;

and generating a user payment portrait data subset corresponding to the preset payment event node based on second payment characteristic data respectively corresponding to the plurality of payment characteristic dimensions.

5. The method of claim 4, wherein the inputting the first historical payment behavior data subset corresponding to the preset payment event node into a pre-trained deep neural network model for recall of payment feature data to obtain second payment feature data corresponding to a plurality of payment feature dimensions respectively comprises:

inputting the first historical payment behavior data subset corresponding to the preset payment event node into a pre-trained deep neural network model for each payment characteristic dimension, and recalling payment characteristic data to obtain a payment behavior characteristic vector corresponding to the payment characteristic dimension; the payment behavior feature vector comprises a first preset number of second payment behavior feature values which are ranked at the top;

and determining second payment characteristic data corresponding to the payment characteristic dimension based on the payment behavior characteristic vector corresponding to the payment characteristic dimension.

6. The method of claim 5, wherein the second payment behavior feature value comprises: the original payment behavior characteristic values of the second preset quantity and the associated payment behavior characteristic values of the third preset quantity; the sum of the second preset number and the third preset number is equal to the first preset number;

the degree of association between the associated payment behavior characteristic value and at least one original payment behavior characteristic value in the second preset number of original payment behavior characteristic values is greater than a preset threshold value.

7. The method of claim 3, wherein prior to obtaining the first set of historical payment behavior data for the first historical period of time for the target user, further comprising:

acquiring a second historical payment behavior data set of the target user in a second historical time period;

and determining a plurality of preset payment event nodes respectively corresponding to at least one payment behavior occurrence cycle type based on the second historical payment behavior data set.

8. The method of claim 7, wherein the determining a plurality of preset payment event nodes respectively corresponding to at least one payment behavior occurrence cycle type based on the second historical payment behavior data set comprises:

determining at least one payment behavior occurrence cycle type involved based on the second historical payment behavior data set;

for each payment behavior occurrence cycle type, determining a plurality of second historical payment behavior data subsets corresponding to the payment behavior occurrence cycle type in the second historical payment behavior data set;

and determining a plurality of preset payment event nodes corresponding to the payment behavior occurrence cycle type based on a plurality of second historical payment behavior data subsets.

9. The method of claim 7, wherein the obtaining, in the set of user payment profile data of the target user, a target payment profile data subset corresponding to a target payment event node to which the pending payment event belongs comprises:

determining the target payment behavior occurrence cycle type to which the payment event to be processed belongs based on the event occurrence time of the payment event to be processed;

determining a target payment event node to which the payment event to be processed belongs in a plurality of preset payment event nodes corresponding to the target payment behavior occurrence cycle type;

and determining a target payment portrait data subset corresponding to the target payment event node in the user payment portrait data set of the target user.

10. An apparatus for identifying a risk of payment, comprising:

the first data acquisition module is used for acquiring a target payment behavior data set of a payment event to be processed of a target user; the target payment behavior data set comprises first payment characteristic data corresponding to a plurality of payment characteristic dimensions generated aiming at the payment event to be processed respectively;

the second data acquisition module is used for acquiring a target payment portrait data subset corresponding to a target payment event node to which the payment event to be processed belongs from the user payment portrait data set of the target user; the user payment portrait data set is obtained by recalling payment characteristic data in advance by utilizing a deep neural network model based on a first historical payment behavior data set of the target user, and the target payment portrait data subset comprises second payment characteristic data corresponding to each payment characteristic dimension under the target payment event node;

and the risk identification result determining module is used for determining a payment risk identification result of the payment event to be processed based on the first payment characteristic data and the second payment characteristic data corresponding to each payment characteristic dimension.

11. The apparatus of claim 10, wherein the first payment characteristic data comprises a first payment behavior characteristic value, and the second payment characteristic data is a payment behavior characteristic vector comprising a plurality of second payment behavior characteristic values; the risk identification result determination module that:

12. The apparatus of claim 10, further comprising: a payment portrait data generation module that:

inputting the first historical payment behavior data set into a pre-trained deep neural network model, and recalling payment characteristic data to obtain a user payment image data subset corresponding to each preset payment event node; the user payment portrait data subset comprises second payment characteristic data corresponding to a plurality of payment characteristic dimensions respectively;

13. An identification device of payment risks comprising:

a processor; and

a memory arranged to store computer executable instructions that, when executed, cause the processor to:

14. A storage medium storing computer-executable instructions that, when executed by a processor, implement a method of: