CN115358772A

CN115358772A - Transaction risk prediction method and device, storage medium and computer equipment

Info

Publication number: CN115358772A
Application number: CN202210925447.4A
Authority: CN
Inventors: 鲍德强; 柳阳
Original assignee: Kangjian Information Technology Shenzhen Co Ltd
Current assignee: Kangjian Information Technology Shenzhen Co Ltd
Priority date: 2022-08-03
Filing date: 2022-08-03
Publication date: 2022-11-18

Abstract

The invention discloses a transaction risk prediction method, a transaction risk prediction device, a storage medium and computer equipment, and relates to the technical field of big data processing. The method comprises the following steps: acquiring user behavior data in a transaction link of a target commodity, wherein the user behavior data comprise wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise a normal state and an abnormal state; obtaining an abnormal trigger value of the target commodity according to the total number of abnormal states of the wind control states of the users on the behavior nodes; and obtaining a transaction risk prediction result of the target commodity according to a comparison result between the abnormal trigger value of the target commodity and a preset trigger threshold value. The method can effectively improve the accuracy of predicting whether the sold commodities have transaction risks on the online platform, and effectively identify whether the current commodities have abnormal risks such as malicious purchase.

Description

Transaction risk prediction method and device, storage medium and computer equipment

Technical Field

The invention relates to the technical field of big data processing, in particular to a transaction risk prediction method, a transaction risk prediction device, a storage medium and computer equipment.

Background

With the explosion of the internet, more and more transactions are migrating from traditional offline traditional channels to online platforms. In order to cultivate the market, each large internet platform also invests a large amount of funds in operation and popularization, and develops a large number of promotion and preferential activities.

But the diversified promotion and preferential activities also provide breeding soil for the black industry of the internet. The malicious purchasing behavior including resource stealing is the most serious damage to each internet e-commerce platform. At present, the users who make malicious purchases adopt more and more cheating means and are more and more abundant, and the users can change the cheating means for many times and bypass the supervision rules to gain profits.

However, the traditional transaction risk prediction method only analyzes the wind control risk data for the current node, and by calculating the request quantity of the current node and taking the request hit condition as an analysis basis, forward analysis is mostly adopted. However, many malicious users do not have abnormality in the current node, and are difficult to identify and perceive the abnormality of commodity activities in the current node, so that the accuracy of transaction risk prediction is low.

Disclosure of Invention

In view of this, the present application provides a transaction risk prediction method, apparatus, storage medium and computer device, and mainly aims to solve the technical problem of low prediction accuracy of the existing transaction risk prediction method.

According to a first aspect of the present invention, there is provided a transaction risk prediction method, the method comprising:

acquiring user behavior data in a transaction link of a target commodity, wherein the user behavior data comprises wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise a normal state and an abnormal state;

obtaining an abnormal trigger value of the target commodity according to the total number of abnormal states of the wind control states of the users on the behavior nodes;

and obtaining a transaction risk prediction result of the target commodity according to a comparison result between the abnormal trigger value of the target commodity and a preset trigger threshold value.

According to a second aspect of the present invention, there is provided a transaction risk prediction apparatus comprising:

the data acquisition module is used for acquiring user behavior data in a transaction link of a target commodity, wherein the user behavior data comprise wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise a normal state and an abnormal state;

the wind control counting module is used for obtaining an abnormal trigger value of the target commodity according to the total number of abnormal wind control states of the users on the behavior nodes;

and the risk judgment module is used for obtaining a transaction risk prediction result of the target commodity according to the abnormal trigger value and a preset trigger threshold value.

According to a third aspect of the present invention, there is provided a storage medium having stored thereon a computer program which, when executed by a processor, implements the above-described transaction risk prediction method.

According to a fourth aspect of the present invention, there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the above-described transaction risk prediction method when executing the program.

According to the transaction risk prediction method, the transaction risk prediction device, the transaction risk prediction medium and the computer equipment, firstly, user behavior data on a plurality of behavior nodes before a target commodity is purchased are traced back, then, the total number of abnormal states of wind control states on all the behavior nodes is obtained according to calculation and is used as a prediction basis of the transaction risk, and finally, whether the target commodity has abnormal risk or not is judged according to the total number of the abnormal states of the wind control states and a preset trigger threshold value. Compared with the existing method for predicting whether the current commodity has transaction risks by respectively judging whether each client has abnormal behaviors based on the personal behaviors of a large number of users, the method can effectively improve the accuracy of predicting whether the commodity sold has the transaction risks by the online platform, and effectively identify whether the current commodity has the abnormal risks such as malicious purchase. Meanwhile, compared with the method for respectively judging whether each customer has abnormal behaviors in the prior art, the method can obviously reduce the calculation amount of commodity transaction risk prediction work and improve the prediction efficiency for identifying whether the commodities have abnormal risks such as malicious purchase.

The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention and do not constitute a limitation of the invention. In the drawings:

fig. 1 is a schematic flow chart illustrating a transaction risk prediction method according to an embodiment of the present invention;

FIG. 2 is a flow chart illustrating another transaction risk prediction method provided by an embodiment of the invention;

fig. 3 is a schematic structural diagram illustrating a transaction risk prediction apparatus according to an embodiment of the present invention.

Detailed Description

The invention will be described in detail hereinafter with reference to the accompanying drawings in conjunction with embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.

At present, a large amount of funds are invested in operation and popularization of each large e-commerce platform, and promotion and preferential activities are developed more and more diversified and complicated. But the complicated and diversified promotion and preferential activities also provide breeding soil for the resource stealing behavior aiming at the e-commerce platform. Resource theft, including malicious purchases, causes significant economic losses to various internet e-commerce platforms. In the face of the current resource stealing means, the traditional transaction risk prediction method only analyzes the wind control risk data aiming at the current node, and adopts forward analysis mostly by calculating the request quantity of the current node and taking the hit condition of the request as an analysis basis. However, many malicious users do not have abnormality in the current node, and are difficult to identify and perceive the abnormality of commodity activities in the current node, so that the accuracy of transaction risk prediction is low.

In view of the above problem, in one embodiment, as shown in fig. 1, a transaction risk prediction method is provided, which is described by taking the method as an example applied to a computer device, and includes the following steps:

101. user behavior data are obtained in a transaction link of a target commodity, wherein the user behavior data comprise wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise normal states and abnormal states.

The user behavior data is generated based on the behavior of the user in the process of purchasing the target commodity and is used for recording the operation behavior of the user. In this embodiment, when a behavior of a user on a behavior node triggers a wind control condition corresponding to the behavior node, a wind control state of the user on the behavior node is calibrated to be an abnormal state, and otherwise, the behavior is a normal state. Specifically, the user behavior data includes operation data of the user at each stage in the transaction link according to the flow sequence, including login request data of the user, login IP data of the user, login device data of the user, recipient information of the user, and the like. The action node refers to a mark used for recording operation data of a user at each stage in a transaction link, and comprises a user login node, a user equipment registration node, a user recipient information filling node and the like.

In this embodiment, all the operation behaviors of the user accessing the internal system of the e-commerce platform need to pass through the gateway, and the gateway sends the received operation behaviors to each behavior node in the transaction link, and each behavior node generates user behavior data corresponding to the cost phase based on the phase of the behavior node in the transaction link and the operation behavior of the current phase. As an example, when a user purchases an e-commerce platform, the user first logs in his account information, the account information and login status information that the user logs in are sent to a gateway of the e-commerce platform, and then the gateway sends the received account information and login status information to a user login node, and the user login node generates user behavior data based on the account information and login status information. Furthermore, each behavior node judges whether the user behavior data generated by each behavior node triggers the wind control condition or not based on the preset wind control condition of the behavior node. And aiming at any behavior node, if the user behavior data triggers a wind control condition, calibrating the wind control state of the user behavior data into an abnormal state. And if the user behavior data does not trigger the wind control condition, calibrating the wind control state of the user behavior data into a normal state. As an example, a transaction link for a target commodity is traced back, and the operation behaviors of all users on the transaction link are analyzed. If the user inputs an incorrect password or inputs incorrect verification information during login operation, the user login node generates user behavior data based on the operation behavior acquired from the gateway, judges that the user behavior data triggers a wind control condition, and then calibrates the user behavior data to be in an abnormal state. If a user replaces a plurality of Internet protocol (ip) addresses or a plurality of hardware addresses to log in during login, the user equipment registration node generates user behavior data based on the operation behavior acquired from the gateway, determines that the user behavior data triggers a wind control condition, and then calibrates the user behavior data to an abnormal state. And if the user inputs a nonexistent address when setting the receiving address, the user recipient information filling node generates user behavior data based on the operation behavior acquired from the gateway, judges that the user behavior data triggers a wind control condition, and then calibrates the user behavior data to be in an abnormal state. Further, if the user does not have the problems of inputting wrong passwords or inputting wrong verification information during login, the user login node generates user behavior data based on the operation behavior acquired from the gateway, judges that the user behavior data does not trigger a wind control condition, and then calibrates the user behavior data to be in a normal state. If the user does not log in by replacing the IP address or the hardware address during logging in, the user equipment registration node generates user behavior data based on the operation behavior acquired from the gateway, judges that the user behavior data does not trigger a wind control condition, and then calibrates the user behavior data to be in a normal state. And if the user inputs a normally existing address when setting a receiving address, the user recipient information filling node generates user behavior data based on the operation behavior acquired from the gateway, judges that the user behavior data does not trigger a wind control condition, and then calibrates the user behavior data to be in a normal state. The method for judging the wind control of the user behavior data on other behavior nodes is similar to that described above, and is not described herein again.

102. And obtaining an abnormal trigger value of the target commodity according to the total number of abnormal states of the wind control states of the users on the behavior nodes.

Specifically, in a transaction link of the target commodity, the number of user behavior data in which the wind control state is calibrated to be an abnormal state on each behavior node is calculated, and then the number of abnormal states of all behavior nodes is accumulated to obtain an abnormal trigger value of the target commodity. In this embodiment, the number of abnormal states of all behavior nodes may also be weighted and accumulated to increase the weight value of the user behavior data of the abnormal state on part of the behavior nodes. When the wind control state of the user behavior data with the high weight value is calibrated to be an abnormal state, the value of the abnormal trigger value can be increased.

103. And obtaining a transaction risk prediction result of the target commodity according to a comparison result between the abnormal trigger value of the target commodity and a preset trigger threshold value.

In this embodiment, a trigger threshold may be set in advance for a target commodity, and it may be determined whether the abnormal trigger value is greater than the trigger threshold. If the abnormal trigger value is larger than the trigger threshold value, judging that the target commodity has a transaction risk; and if the abnormal trigger value is not greater than the trigger threshold value, judging that the target commodity has no transaction risk. The trigger threshold may be set based on actual conditions or historical data, or may be obtained by correcting a preset number of basic users based on an actual number of users.

According to the transaction risk prediction method provided by the embodiment, firstly, user behavior data of a target commodity on a plurality of behavior nodes in a transaction link are traced back, and then the total number of wind control states on all the behavior nodes, which are abnormal states, is obtained according to calculation and is used as a prediction basis of transaction risks. And finally, judging whether the target commodity has abnormal risks or not based on the total amount of the user behavior data with the wind control state as the abnormal state and a preset trigger threshold value. The invention can effectively improve the accuracy of predicting whether the sold commodities have transaction risks on the online platform and effectively identify whether the current commodities have abnormal risks such as malicious purchase and the like. Meanwhile, compared with the method for respectively judging whether each customer has abnormal behaviors in the prior art, the method can obviously reduce the calculation amount of commodity transaction risk prediction work and improve the prediction efficiency for identifying whether the commodities have abnormal risks such as malicious purchase.

In one embodiment, before step 102, the transaction risk prediction method further includes: firstly, coding the plurality of behavior nodes to obtain the node code of each behavior node. In this embodiment, because the amount of the acquired data is too large, the user behavior information needs to be transcoded before being stored, so as to reduce the storage space of the data. Specifically, the behavior nodes are transcoded, and each behavior node is transcoded into a specific node code. By way of example, the user login node may be encoded as a, the user device registration node may be encoded as B, and the user recipient information filling node may be encoded as C. And then, transcoding the user behavior data according to the node code of each behavior node to obtain the behavior code of each user on each behavior node. And the behavior code of each user on each behavior node corresponds to one wind control state. Specifically, each transcoded behavior node transcodes the user behavior data generated by the behavior node based on the operation behavior of the user, and a behavior code of the user on each behavior node is obtained. As an example, if the node code of the current behavior node is a, the user behavior data generated by the current behavior node is uniformly transcoded into the user behavior data of behavior code a. Further, each behavior code will also have the same wind-controlled state as the user behavior data before transcoding. By utilizing the technical scheme provided by the application, the space occupied by the behavior code after transcoding can be obviously smaller than the user behavior data before transcoding, the space required by data storage can be effectively reduced, and the problem of cost increase caused by large-scale data calculation is avoided. Furthermore, the user behavior data generated by each behavior node is transcoded into the behavior codes, and the behavior codes are the same as the node codes, so that the behavior codes corresponding to the same behavior node can be the same, and the calculation amount for classifying the behavior codes in the later period is reduced.

In this embodiment, the wind control state condition of the target commodity may also be stored, and the number of times of triggering the abnormal state by the wind control state of each behavior code group corresponding to the target commodity and the storage time are stored.

In one embodiment, the step 102 may be implemented by:

201. and dividing the same behavior codes in the user behavior data into the same behavior code group.

In this embodiment, because the user behavior data generated by the same transcoded behavior node all have the same behavior code, the same behavior code may be divided into the same behavior code group.

As an example, if the node code after transcoding the user login node code is a, the user behavior data generated by the behavior node is transcoded into the user behavior data of the behavior code a in a unified manner. All the behavior codes A are divided into the same behavior code group, so that the calculation pressure in the later accumulation process can be reduced.

202. And acquiring a wind control state corresponding to each behavior code in each behavior code group, and counting the number of abnormal states of the wind control state to obtain the abnormal times of each behavior code group.

Specifically, each behavior code corresponds to one of the wind control states, so that the sum of the behavior codes with abnormal states in the behavior code group is counted, and the abnormal times of each behavior code group can be obtained.

203. And accumulating the abnormal times of each behavior code group to obtain an abnormal trigger value of the target commodity.

In this embodiment, the same behavior codes are collected into the same behavior code group, then the abnormal times in each behavior code group are accumulated to obtain the abnormal times of each behavior code group, and then the abnormal times of each behavior code group are accumulated to obtain the abnormal trigger value of the target commodity. The behavior codes are collected to obtain behavior code groups, the abnormal times of each behavior code group are respectively judged, the abnormal times do not need to be calculated from massive mixed and stored behavior codes, the calculated amount of the calculation work of the trigger value can be reduced, and the calculation pressure is reduced.

In one embodiment, each of the behavior code groups corresponds to a preset weight modification value. Further, the step 203 may be implemented by: firstly, obtaining the abnormal times correction value of each behavior coding group according to the product of the abnormal times of each behavior coding group and the weight correction value of the behavior coding group. Specifically, for key cheating behaviors frequently used by a user who steals resources, user behavior data corresponding to the key cheating behaviors are selected, and the weight correction value of a behavior code group where the user behavior data is located is improved, so that the behavior code group can obtain a higher abnormal frequency correction value. And then, obtaining an abnormal trigger value of the target commodity according to the sum of the abnormal time correction values of each behavior code group. In this embodiment, different weights are assigned to the behavior code groups corresponding to different user behavior data, so that the behavior code group corresponding to the key cheating behavior obtains a higher abnormal number correction value, and the abnormal trigger value of the target product is further increased. Furthermore, the probability that the abnormal trigger value corresponding to the target commodity subjected to resource stealing by key cheating is larger than the trigger threshold value is increased, and the prediction accuracy for judging whether the target commodity has a transaction risk is improved.

In one embodiment, the trigger threshold may be obtained by: firstly, the number of users corresponding to the user behavior data is obtained. Specifically, the number of users who generate the user behavior data may be obtained based on the transaction link of the target product. Then, calculating a difference value between the user number and a preset basic user number, and obtaining a corrected user number according to a product of the difference value and a preset correction percentage. The number of basic users can be set according to the historical purchasing number of the commodity, and the correction percentage can be set based on actual conditions and requirements. When the number of users is less than the base number of users, the correction percentage may be increased accordingly. As an example, if the purchase amount of a certain product in the history unit time is 1000, the base user number of the product may be set to 1000. Further, if the obtained number of users is 5000, the difference between the number of users and the base number of users is 4000. If the correction percentage is 10%, the number of corrected users is 400. And finally, obtaining the trigger threshold according to the sum of the number of the corrected users and the number of the basic users. As an example, if the number of modified users is 400 and the number of base users is 1000, the trigger threshold is 1400. Further, if the obtained number of users is 500 and the number of basic users is set to 1000, the difference between the number of users and the number of basic users is 500. If the correction percentage is 160%, the number of corrected users is-800. At this point, the trigger threshold is calculated to be 200. According to the method, the preset basic user number is corrected according to the difference value between the actual user number and the basic user number, and when the difference between the actual user number and the basic user number is large, the basic user number is corrected to obtain the trigger threshold. And correspondingly increasing the trigger threshold when the actual number of users is large, and correspondingly decreasing the trigger threshold when the actual number of users is small. The method avoids the error prediction of judging the target commodity to have the transaction risk based on a smaller abnormal trigger value caused by the fact that the actual number of users is larger and the trigger threshold value is still in a lower value. Or the actual number of users is small, and the trigger threshold is still at a high value, resulting in a wrong prediction that the target product is determined not to have a transaction risk even based on a large abnormal trigger value. The elasticity of the prediction work is increased, and the prediction accuracy for judging whether the target commodity has the transaction risk can be improved.

In one embodiment, as shown in fig. 2, after step 103, the transaction risk prediction method further comprises the steps of:

104. and if the target commodity has a transaction risk, acquiring user information for purchasing the target commodity within a preset time range.

In this embodiment, the time range may be set based on actual conditions and requirements. Further, the user information can be determined based on the recipient information collected by the user recipient information filling node. If the address information and the contact information in the plurality of pieces of recipient information are the same, the recipient information can be determined as the same piece of user information.

Specifically, the time range may be set to one day or one week. Further, the determination may be based on the average mailing time from shipment to receipt of the target item in the history. If the time range is set to be a value smaller than the average mailing time, the e-commerce platform is ensured to recall the express delivery before the resource-stealing user receives the express delivery.

105. And generating an early warning mail according to the user information, and sending the early warning mail.

The early warning mail can contain user information of all purchasing behaviors and user information corresponding to user behavior data of which all wind control states are abnormal states.

Specifically, after the early warning mail is received, the transaction condition of the target commodity can be repaired, and a right-maintaining action can be performed based on the user information of the early warning mail. Furthermore, a reminding short message can be sent to a manager of the e-commerce platform or a reminding telephone can be dialed, so that the manager can know that the target commodity has a transaction risk at the first time.

According to the transaction risk prediction method provided by the embodiment, the abnormal trigger value can be obtained by backtracking the user behavior data of the target commodity on the plurality of behavior nodes before the purchasing node and calculating the quantity of the user behavior data with the abnormal state based on the wind control state. And then, judging whether the target commodity has abnormal risks or not according to the magnitude relation between the abnormal trigger value and the trigger threshold value. The invention can effectively improve the accuracy of predicting whether the sold commodities have transaction risks on the online platform and effectively identify whether the current commodities have abnormal risks such as malicious purchase and the like. Meanwhile, compared with the method for respectively judging whether each customer has abnormal behaviors in the prior art, the method can obviously reduce the calculation amount of commodity transaction risk prediction work and improve the prediction efficiency for identifying whether the commodities have abnormal risks such as malicious purchase.

Further, as a specific implementation of the method shown in fig. 1 and fig. 2, the embodiment provides a transaction risk prediction apparatus, as shown in fig. 3, the apparatus includes: a data acquisition module 31, a wind control counting module 32 and a risk determination module 33.

The data acquisition module 31 may be configured to acquire user behavior data in a transaction link of a target commodity, where the user behavior data includes a plurality of wind control states of users on a plurality of behavior nodes, and the wind control states include a normal state and an abnormal state;

the wind control counting module 32 is configured to obtain an abnormal trigger value of the target commodity according to the total number of abnormal states of the wind control states of the users at the behavior nodes;

the risk determination module 33 may be configured to obtain a transaction risk prediction result of the target product according to a comparison result between the abnormal trigger value of the target product and a preset trigger threshold.

In a specific application scenario, the data obtaining module 31 may be further configured to encode the plurality of behavior nodes to obtain a node code of each behavior node; and transcoding the user behavior data according to the node code of each behavior node to obtain the behavior code of each user on each behavior node, wherein the behavior code of each user on each behavior node corresponds to one wind control state.

In a specific application scenario, the wind control counting module 32 may be specifically configured to divide the same behavior codes in the user behavior data into the same behavior code group; acquiring a wind control state corresponding to each behavior code in each behavior code group, and counting the number of abnormal states of the wind control state to obtain the abnormal times of each behavior code group; and accumulating the abnormal times of each behavior code group to obtain an abnormal trigger value of the target commodity.

In a specific application scene, each behavior coding group corresponds to a preset weight modification value; the wind control counting module 32 is further configured to obtain a corrected value of the abnormal times of each behavior coding group according to a product of the abnormal times of each behavior coding group and the corrected value of the weight of the behavior coding group; and obtaining an abnormal trigger value of the target commodity according to the sum of the abnormal times correction values of each behavior coding group.

In a specific application scenario, the risk determining module 33 may be further configured to determine whether the abnormal trigger value is greater than the trigger threshold value; if the abnormal trigger value is larger than the trigger threshold value, judging that the target commodity has a transaction risk; and if the abnormal trigger value is not greater than the trigger threshold value, judging that the target commodity has no transaction risk.

In a specific application scenario, the risk determination module 33 may be further configured to, if the target product has a transaction risk, obtain user information for purchasing the target product within a preset time range; and generating an early warning mail according to the user information, and sending the early warning mail.

It should be noted that other corresponding descriptions of the functional units related to the transaction risk prediction apparatus provided in this embodiment may refer to the corresponding descriptions in fig. 1 and fig. 2, and are not repeated herein.

Based on the methods shown in fig. 1 and fig. 2, correspondingly, the present embodiment further provides a storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the transaction risk prediction method shown in fig. 1 and fig. 2.

Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, and the software product to be identified may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, or the like), and include several instructions to enable a computer device (which may be a personal computer, a server, or a network device, or the like) to execute the method described in the implementation scenarios of the present application.

Based on the method shown in fig. 1 and fig. 2 and the embodiment of the transaction risk prediction apparatus shown in fig. 3, in order to achieve the above object, the embodiment further provides an entity device for predicting transaction risk, which may specifically be a personal computer, a server, a smart phone, a tablet computer, a smart watch, or other network devices, and the entity device includes a storage medium and a processor; a storage medium for storing a computer program; a processor for executing the computer program to implement the above-mentioned methods as shown in fig. 1 and fig. 2.

Optionally, the entity device may further include a user interface, a network interface, a camera, a Radio Frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, and the like. The user interface may include a Display screen (Display), an input unit such as a keypad (Keyboard), etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface), etc.

Those skilled in the art will appreciate that the physical device structure of the transaction risk prediction provided in the present embodiment does not constitute a limitation to the physical device, and may include more or less components, or combine some components, or arrange different components.

The storage medium can also comprise an operating system and a network communication module. The operating system is a program for managing the hardware of the entity device and the software resources to be identified, and supports the running of the information processing program and other software and/or programs to be identified. The network communication module is used for realizing communication among components in the storage medium and other hardware and software in the information processing entity equipment.

Through the above description of the embodiments, those skilled in the art will clearly understand that the present application can be implemented by software plus a necessary general hardware platform, and can also be implemented by hardware. By applying the technical scheme of the application, firstly, user behavior data are obtained in a transaction link of a target commodity, wherein the user behavior data comprise wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise a normal state and an abnormal state; then, obtaining an abnormal trigger value of the target commodity according to the total number of abnormal states of the wind control states of the users on the behavior nodes; and finally, obtaining a transaction risk prediction result of the target commodity according to a comparison result between the abnormal trigger value of the target commodity and a preset trigger threshold value. Compared with the prior art, the method can obviously reduce the calculation amount of commodity transaction risk prediction work, and improve the prediction efficiency of identifying whether the commodity has abnormal risks such as malicious purchase.

Those skilled in the art will appreciate that the drawings are merely schematic representations of preferred embodiments and that the blocks or flowchart illustrations are not necessary to practice the present application. Those skilled in the art can understand that the modules in the device in the implementation scenario may be distributed in the device in the implementation scenario according to the implementation scenario description, and may also be located in one or more devices different from the present implementation scenario with corresponding changes. The modules of the implementation scenario may be combined into one module, or may be further split into a plurality of sub-modules.

The above application serial numbers are for description purposes only and do not represent the superiority or inferiority of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of the present application, but the present application is not limited thereto, and any variations that can be made by those skilled in the art are intended to fall within the scope of the present application.

Claims

1. A method for predicting transaction risk, the method comprising:

acquiring user behavior data in a transaction link of a target commodity, wherein the user behavior data comprise wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise a normal state and an abnormal state;

2. The method according to claim 1, wherein before the obtaining of the abnormal trigger value of the target commodity according to the total number of abnormal states of the wind control states of the users on the behavior nodes, the method further comprises:

coding the plurality of behavior nodes to obtain a node code of each behavior node;

and transcoding the user behavior data according to the node code of each behavior node to obtain the behavior code of each user on each behavior node, wherein the behavior code of each user on each behavior node corresponds to one wind control state.

3. The method according to claim 2, wherein the obtaining of the abnormal trigger value of the target commodity according to the total number of abnormal states of the wind-controlled states of the users at the behavior nodes comprises:

dividing the same behavior codes in the user behavior data into the same behavior code group;

acquiring a wind control state corresponding to each behavior code in each behavior code group, and counting the number of abnormal states of the wind control state to obtain the abnormal times of each behavior code group;

and accumulating the abnormal times of each behavior code group to obtain an abnormal trigger value of the target commodity.

4. The method according to claim 3, wherein each of the behavior code groups corresponds to a predetermined weight modification value;

the step of accumulating the abnormal times of each behavior code group to obtain the abnormal trigger value of the target commodity comprises the following steps:

obtaining the abnormal time correction value of each behavior coding group according to the product of the abnormal time of each behavior coding group and the weight correction value of the behavior coding group;

and obtaining the abnormal trigger value of the target commodity according to the sum of the abnormal times correction value of each behavior code group.

5. The method as claimed in claim 1, wherein the method for obtaining the trigger threshold comprises:

acquiring the number of users corresponding to the user behavior data;

calculating a difference value between the user number and a preset basic user number, and obtaining a corrected user number according to a product of the difference value and a preset correction percentage;

and obtaining the trigger threshold according to the sum of the number of the corrected users and the number of the basic users.

6. The method according to claim 1 or 5, wherein obtaining a transaction risk prediction result of the target commodity according to a comparison result between the abnormal trigger value of the target commodity and a preset trigger threshold value comprises:

judging whether the abnormal trigger value is larger than the trigger threshold value or not;

if the abnormal trigger value is larger than the trigger threshold value, judging that the target commodity has a transaction risk;

and if the abnormal trigger value is not greater than the trigger threshold value, judging that the target commodity has no transaction risk.

7. The method according to claim 6, wherein after obtaining the predicted transaction risk result of the target product according to the comparison result between the abnormal trigger value of the target product and the preset trigger threshold, the method further comprises:

if the target commodity has a transaction risk, acquiring user information for purchasing the target commodity within a preset time range;

and generating an early warning mail according to the user information, and sending the early warning mail.

8. A transaction risk prediction apparatus, the apparatus comprising:

the data acquisition module is used for acquiring user behavior data in a transaction link of a target commodity, wherein the user behavior data comprises wind control states of a plurality of users on a plurality of behavior nodes, and the wind control states comprise a normal state and an abnormal state;

and the risk judgment module is used for obtaining a transaction risk prediction result of the target commodity according to a comparison result between the abnormal trigger value of the target commodity and a preset trigger threshold value.

9. A storage medium on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.

10. A computer arrangement comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the computer program realizes the steps of the method of any one of claims 1 to 7 when executed by the processor.