CN117172532A

CN117172532A - Wind control service providing method and device

Info

Publication number: CN117172532A
Application number: CN202311009369.4A
Authority: CN
Inventors: 宋佳; 石翼
Original assignee: AlipayCom Co ltd
Current assignee: AlipayCom Co ltd
Priority date: 2023-08-10
Filing date: 2023-08-10
Publication date: 2023-12-05

Abstract

The embodiment of the specification discloses a wind control service providing method, which is applied to a wind control platform and comprises the following steps: acquiring historical operation behavior data of a target merchant in a preset time period; determining the current safe and reliable score of the target merchant based on the historical operation behavior data; determining a low-risk merchant and a high-risk merchant based on the safe trusted score and a preset control level of the target merchant; initiating a preset autonomous risk management interaction service to the low-risk merchant based on the risk category of the low-risk merchant; after the low-risk merchant completes the autonomous risk management interactive service, updating the safe and trusted score of the low-risk merchant; and determining a management and control strategy for the high-risk merchant through a pre-trained strategy recommendation model based on the historical operation behavior data of the high-risk merchant. Accordingly, the invention discloses a wind control service providing device.

Description

Wind control service providing method and device

Technical Field

The present invention relates to the field of computer technologies, and in particular, to a method and an apparatus for providing a wind control service.

Background

With the continuous upgrading of crime and attack and defense technologies in the black ash industry, the layering identification difficulty of black ash merchants and normal operation merchants is continuously increased, and the original wind control scheme is upgraded for improving the wind control capability of each platform. And excessive reinforcement of wind control force can exist to recognize normal transaction as the possibility of risk action, leads to the unreasonable condition of management and control means of propelling movement, influences user experience.

Disclosure of Invention

The invention aims to provide a wind control service providing method which can realize balance between risk control and user experience and scientific and reuse the wind control experience of users.

According to the above object, an embodiment of the present disclosure provides a method for providing a wind control service, which is applied to a wind control platform, including:

acquiring historical operation behavior data of a target merchant in a preset time period;

inputting the historical operation behavior data into a pre-trained safe credible score evaluation model, and determining the current safe credible score of the target merchant based on the evaluation result of the safe credible score evaluation model;

determining a low-risk merchant and a high-risk merchant based on the safe trusted score and a preset control level of the target merchant;

initiating a preset autonomous risk management interaction service to the low-risk merchant based on the risk category of the low-risk merchant; after the low-risk merchant completes the autonomous risk management interactive service, updating the safe and trusted score of the low-risk merchant;

and determining a management and control strategy for the high-risk merchant through a pre-trained strategy recommendation model based on the historical operation behavior data of the high-risk merchant.

According to the wind control service providing method provided by the embodiment of the specification, the risk condition of the target merchant is determined by evaluating the safe and reliable score, education prompt and front-end problem solving are carried out on the target merchant with the existing risk trend, the target merchant is promoted to actively participate in own risk management, and risk behaviors are avoided; and performing scientific control on the target merchant with risk behaviors.

Further, in some embodiments, before determining the low-risk merchant and the high-risk merchant based on the secure trust score, further comprising:

determining the industry category of the target merchant based on the basic information and the business behavior data of the target merchant;

determining a risk-free merchant based on the industry category;

and eliminating the risk-free merchant from the target merchant.

Further, in some embodiments, the safe trusted score evaluation model is pre-trained in the following manner, including:

constructing a training sample; the training samples comprise positive samples and negative samples, the positive samples comprise different types of positive business behavior data, and the negative samples comprise different types of negative business behavior data;

setting corresponding safe trusted points as labels based on the types of the positive operation behavior data and the negative operation behavior data;

Training a pre-constructed classification model by using the training sample and the label until the safe and reliable score evaluation model meeting the preset condition is obtained.

Further, in some embodiments, after the low-risk merchant completes the autonomous risk management interaction service, updating the security trusted score of the low-risk merchant specifically includes:

acquiring business behavior data of the low-risk merchant in the process of completing the autonomous risk management interactive service;

inputting the operation behavior data into the safe and reliable score evaluation model;

and updating the current safe trusted score of the low-risk merchant based on the evaluation result of the safe trusted score evaluation model.

Further, in some embodiments, the policy recommendation model is pre-trained in the following manner:

determining a management and control strategy set based on the history management and control data;

for each management policy in the management policy set, collecting first operation behavior data of a corresponding merchant before the management policy is implemented, and collecting second operation behavior data of the corresponding merchant after the management policy is implemented and complaint data of the corresponding merchant on the management policy;

Determining a control effect score of the control strategy based on the second business behavior data;

determining environmental information based on the second business activity data;

determining an incentive signal based on the management and control effect score and the complaint data;

and performing reinforcement learning based on the environment information and the excitation signal until the strategy recommendation model meeting the preset condition is obtained.

Further, in some embodiments, further comprising:

and after the management and control strategy is implemented for the high-risk merchant, feeding back management and control prompt information to the high-risk merchant, and embedding complaint channel entrance links in the display page of the management and control prompt information.

Still further, in some embodiments, the regulatory prompt is used to describe a business risk behavior category of the high-risk merchant; and the complaint channel entrance links embedded in the display page are used for pointing to the complaint channels corresponding to the business risk behavior categories.

Still further, in some embodiments, further comprising:

acquiring complaint materials submitted by the high-risk merchants through the complaint channels;

pre-auditing the materials meeting the pre-auditing conditions in the application materials based on a preset auditing model;

If the pre-verification result is not passed, the non-passed verification content is fed back to the high-risk merchant for modification;

and if the pre-verification result is passed, transferring the complaint material to a manual verification node for continuous verification.

Further, in some embodiments, further comprising:

after executing the control strategy on the high-risk merchant, acquiring the control behavior data of the high-risk merchant aiming at the control strategy through a node preset in a control channel;

determining a solution behavior path of the high-risk merchant based on the node where the solution behavior data occurs;

and determining user experience barrier points on the solution control behavior path based on the solution control behavior path and a preset solution control path.

Another object of the present invention is to provide a wind control service providing apparatus, which can optimize user experience, and achieve a balance between risk control and user experience.

According to the above object, an embodiment of the present specification proposes a wind control service providing apparatus including:

the data acquisition unit is configured to acquire historical operation behavior data of a target merchant in a preset time period;

The safe credibility module is configured to input the historical operation behavior data into a pre-trained safe credibility evaluation model, and determine the current safe credibility of the target merchant based on the evaluation result of the safe credibility evaluation model; the method comprises the steps that a target merchant receives a preset autonomous risk management interaction service, and the safety and credibility of the target merchant are updated in response to the completion of the preset autonomous risk management interaction service by the target merchant;

the wind control strategy generation module is configured to determine a low-risk user and a high-risk user based on the safe and reliable score; initiating a preset autonomous risk management interaction task based on the risk category of the low-risk merchant; determining a management and control strategy for the high-risk merchant through a strategy recommendation model based on the historical operation behavior data of the high-risk merchant, and initiating a high-risk wind control task;

a wind control engine configured to provide a preset autonomous risk management interaction service to the low-risk merchant in response to the autonomous risk management interaction task; and responsive to the high risk wind control task, enforcing the governance policy for the high risk user.

The embodiment of the specification also provides an electronic device, including:

a processor;

A memory for storing the processor-executable instructions;

wherein the processor implements the method as described in any one of the wind-controlled service provision methods above by executing the executable instructions.

The present description embodiments also provide a computer-readable storage medium storing a computer program which, when executed by a processor, implements a method as set forth in any one of the wind-controlled service providing methods above.

The wind control service providing method has the advantages that the risk condition of the target merchant is determined by evaluating the safe and reliable score, education prompts and front-end problem solving are carried out on the target merchant with the existing risk trend, the target merchant is promoted to actively participate in self risk management, and risk behaviors are avoided; the method has the advantages that scientific management and control are carried out on target merchants with risk behaviors, management and control prompt information is directly and effectively transmitted to the risk merchants, and a complaint channel and a complaint material pre-examination link are provided, so that the understandability of merchant management and control information and the fluency of complaint links are improved, different complaint schemes are recommended based on merchant values and management and control means, and the efficiency of post-air-control service is improved. In addition, after the management and control strategy is executed for the high-risk merchant, the solution control behavior path is obtained and analyzed, so that the wind control experience barrier point of the target merchant can be restored, and evaluation feedback data operation aiming at the management and control strategy is realized.

The wind control service providing device disclosed by the embodiment of the specification has the same beneficial effects.

Drawings

In order to more clearly illustrate the embodiments of the present description or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are some embodiments of the present description, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 schematically illustrates a structural diagram of a wind control service system according to one or more embodiments of the present disclosure.

Fig. 2 schematically illustrates a structural schematic diagram of a wind-controlled platform according to one or more embodiments of the present disclosure.

Fig. 3 schematically illustrates a flow diagram of a method for providing a wind control service according to one or more embodiments of the present disclosure, in an implementation manner.

FIG. 4 schematically illustrates a flow diagram of a strategy recommendation model training method according to one or more embodiments of the present disclosure, under one implementation.

FIG. 5 schematically illustrates a flow diagram of an audit method in a method of providing a wind controlled service according to one or more embodiments of the present disclosure, in one embodiment.

Fig. 6 is a block diagram schematically illustrating a structure of an air control service providing apparatus according to one or more embodiments of the present disclosure.

Fig. 7 exemplarily shows a block diagram of an electronic device provided in an embodiment of the present specification.

Detailed Description

First, it will be understood by those skilled in the art that the terminology used in the embodiments of the present invention is for the purpose of describing particular embodiments only, and is not intended to be limiting of the invention. As used in this application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Along with the progress of science and technology and the popularization of the Internet, under the trend of gradually expanding the scale of merchants in various industries, the platform is responsible for striking illegal behaviors and maintaining the order of the platform by improving the wind control capability of the platform where the merchants are located and implementing a proper control mode. For risk control schemes, it is important to seek a balance of risk control and user experience. If the wind control is excessively enhanced, for example, a multiple user identity review link is added, the user experience is affected, the payment desire of the user is reduced, the payment success rate is reduced, and the development of the platform is affected.

In the scenes of merchant collection and the like, as the black-gray merchants are continuously updated by utilizing the technology of benefit seeking by an abnormal means, the difficulty of distinguishing the black-gray merchants from normal business merchants is increased, the condition that management and control are unreasonable due to inaccurate risk judgment is easily generated by determining a risk management and control method through automatically identifying merchant risk behaviors, the management and control of the normal merchants are influenced, the determined management and control method is more mandatory, the applicability of the specific risk merchants is lacked, and the user experience is greatly discounted.

On the other hand, after the merchant is managed and controlled, jump and breakpoint conditions exist between complaint page guidance displayed by the platform and the service area capable of actually processing the problem, and a more direct and smoother post-service scheme is needed; the wind control information informed in the service battlefield needs to be more detailed, such as providing a cause of risk or limiting the use of merchants in the management process. In addition, taking a merchant collection scene as an example, because of lack of a data exchange process with a payment system, a risk page appears on a payer terminal, so that a merchant cannot receive risk prompt information in time. Therefore, the post-service scheme after risk management for merchants is also required to be further improved.

In view of this, one or more embodiments of the present disclosure propose a method for providing a wind control service, which optimizes aspects of wind control decision, active touch before control, real-time guidance after control, service area construction, etc., so as to balance between risk control and user experience, timely inform a risk merchant of a control reason while avoiding risk in advance, and provide an effective complaint entrance for the risk merchant.

In order to make the technical solutions in the present specification better understood by those skilled in the art, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only some embodiments of the present specification, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.

It should be noted that: in other embodiments, the steps of the corresponding method are not necessarily performed in the order shown and described in this specification. In some other embodiments, the method may include more or fewer steps than described in this specification. Furthermore, individual steps described in this specification, in other embodiments, may be described as being split into multiple steps; while various steps described in this specification may be combined into a single step in other embodiments.

One or more embodiments of the present application provide a wind control service providing method. Referring to fig. 1, fig. 1 illustrates a merchant collection scene as an example, which shows a wind control service system that may be used to deploy a wind control platform to implement the wind control service providing method. It should be noted that, the method for providing the wind control service according to one or more embodiments of the present application may be implemented by the wind control service system shown in fig. 1, but is not limited to the wind control service system.

As shown in fig. 1, the wind-controlled service system includes a wind-controlled platform 10 and a payee 20, wherein the payee 20 includes a payee terminal 22 and a server 24. The collecting terminal 22 is connected to the air control platform 10 and the server 24 through communication links, which may be a wired network or a wireless network. For example, the checkout terminal 22 may establish a communication connection with the air control platform 10 and/or the server 24 using WIFI, bluetooth, infrared, etc. communication. Alternatively, the collecting terminal 22 may also establish a communication connection with the wind control platform 10 and/or the server 24 through a mobile network, where the network system of the mobile network may be any of 2G (GSM), 2.5G (GPRS), 3G (WCDMA, TD-SCDMA, CDMA2000, UTMS), 4G (LTE), 4g+ (lte+), wiMax, and the like.

As shown in fig. 2, the wind control platform 10 includes a wind control decision layer 12, a real-time interaction layer 14, a post-service layer 16 and an evaluation feedback layer 18, which are respectively used for evaluating the risk of the business activity of the merchant of the payee, identifying the risk behavior and implementing risk management by optimizing the wind control decision, autonomous risk management before management, construction of the post-management service array and quantitative evaluation feedback mechanism. The risk platform 10 can be implemented in a wind control service system, and communicates management and control prompt information in real time by communicating with the payee 20, so that a target merchant with risk can discover and remove the risk in time. In this embodiment, the implementation form of the wind control platform 10 is not limited, for example, the wind control platform 10 may be deployed in a single server, or may be deployed in a server cluster formed by a plurality of servers, and the wind control platform 10 may also be deployed in a cloud server, which is also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system.

The payee 20 is comprised of a payee terminal 22 and a corresponding server 24. The collection terminal 22 may be a mobile terminal such as a mobile phone, a notebook, a tablet computer, etc., in which a front-end app provided by a platform is installed for conducting transactions such as payment and collection, and the related technical carriers include, but are not limited to, near Field Communication (NFC), WIFI, 3G/4G/5G, POS card swiping technology, two-dimensional code scanning technology, bar code scanning technology, bluetooth, infrared, short Message (SMS), multimedia Message (MMS), etc.

The server 24 corresponding to the collection terminal 22 is used for recording and storing data generated during the transaction of the target merchant, and transmitting the data to the wind control platform when necessary. The server 24 may be any device, apparatus, platform, cluster of devices having computing and processing capabilities, either as a single server or as a cluster of servers.

When the wind control service system implements the wind control service providing method described in this embodiment, the wind control decision layer 12 in the wind control platform 10 obtains the historical operation behavior data of the target merchant in the preset time period through the server 24 of the payee 20, and evaluates the risk of the target merchant; if the target merchant is determined to be a low-risk merchant, the real-time interaction layer 14 of the pneumatic control platform 10 provides autonomous risk management interaction services to enable the target merchant to autonomously reduce the risk; if the information is determined to be a high-risk merchant, the wind control platform 10 limits the related functions of the target merchant by using the management and control strategy decided by the wind control decision layer 12, for example, the wind control platform 10 limits the collection function of the target merchant, not only transmits the payment interception error reporting information to the user terminal of the payer, but also transmits the payment interception error reporting information to the server 24 of the payee 20 through the post service layer 16, and then the server 24 transmits the payment interception error reporting information to the collection terminal 22, thereby opening up the risk platform and the payment system, and enabling the target merchant serving as the payee 20 to receive the risk management and control prompt in real time so as to enter the complaint link and the audit link of the post service layer 16 to release the risk. In the risk relieving process, the evaluation feedback layer 18 in the wind control platform 10 builds a user path line large graph by burying points on the solution control path of the target merchant so as to grasp the blocking situation encountered in the solution control link and quantitatively evaluate and feed back the management and control strategy decided by the wind control decision layer 12.

In response to the above-described wind control service system, in some embodiments, the present specification provides a wind control service providing method applied to a wind control platform. Referring to fig. 3, the method includes the following steps:

s100: and acquiring historical operation behavior data of the target merchant in a preset time period.

In order to obtain historical business behavior data of a more representative target merchant, so that the historical business behavior data can describe as many merchant business conditions as possible, the interval of the preset time period is not too short, and meanwhile, the storage cost and the operation rate are considered, and a proper preset time period is selected.

The historical business behavior data of the target merchant can reflect information such as the collection behavior, business category, buyer feedback and the like of the merchant, is favorable for restoring the real business scene of the target merchant, and can evaluate the risk and credibility of the target merchant more accurately based on multi-element and objective data. The data reflecting the collection behavior of the target merchant includes, but is not limited to, collection product information, collection mode, single collection amount, collection frequency, daily business, etc., and the data related to the collection behavior can be obtained from the server corresponding to the collection terminal. In addition, the historical business behavior data of the target merchant can also comprise buyer LBS aggregation conditions, and can reflect the main business place and audience geographic position distribution; the abnormal operation sequence of the merchant can record abnormal behaviors such as abnormal large-amount transactions, unqualified inspection quality and the like through monitoring the daily business behaviors of the merchant.

S102: and inputting the historical operation behavior data into a pre-trained safe credible score evaluation model, and determining the current safe credible score of the target merchant based on the evaluation result of the safe credible score evaluation model.

The security and credibility of the operation activities of the target merchants can be reflected by the security and credibility of the operation activities of the target merchants, the higher the security and credibility is, the lower the possibility of risk events such as illegal actions and the like of the target merchants is, the lower the possibility of the target merchants being black-gray merchants is, and the transaction environments of the payee and the payer are relatively safer; if the security and credibility score is low, the target merchant has a strong risk tendency, even the risk caused by illegal actions is identified, and in order to avoid the loss of resources such as funds and personal information of both transaction parties, measures need to be taken as soon as possible to reduce the risk or a risk management and control strategy needs to be implemented immediately.

The historical business behavior data of the target merchant can be divided into positive business behavior data and negative business behavior data, and the information which can be obtained by analyzing the negative business behavior data of the merchant includes, but is not limited to, merchant business conditions, abnormal behaviors and external evaluations. The merchant operation status information may include whether a transaction dispute is generated, such as whether a collection amount is inconsistent with a contract, whether repeated collection is performed, whether goods or services are seriously inconsistent with descriptions, so that consumer complaints are caused, and the like, and the safe and reliable score evaluation model evaluates safe and reliable scores according to the severity of the dispute so as to identify the condition that a black gray merchant obtains illegal benefits by doing hands and feet in a transaction manner; the method can also comprise the steps of judging whether the commodity operated by the target merchant belongs to a forbidden commodity or a limited commodity, judging whether the limited commodity is sold according to relevant regulations, judging whether the target merchant has false marketing phenomenon in the propaganda process, judging whether the target merchant has dishonest conditions such as non-regulated performance after signing with the platform; for another example, the bottom line risk of the target merchant is developed and examined, and related consumer complaints are obtained, and the safe and reliable score evaluation model estimates the risk bearing capacity of the target merchant according to the bottom line risk, so that the safe and reliable score of the target merchant is estimated.

On the other hand, the security and credibility score can be evaluated according to the abnormal behavior of the merchant, such as detecting whether abnormal transaction (such as large-amount transaction) exists in the transaction scene of the merchant and the frequency and the amount of the abnormal transaction; the quality and the checking result of the qualification materials submitted by the target merchant during annual inspection and/or patrol inspection can influence the checking result by the authenticity and the validity of the qualification materials; and then, capturing the behavior sequence of the target merchant through a related algorithm, for example, capturing the behavior of the target merchant in the links of goods incoming, processing, propaganda, sales and the like, and further, recognizing abnormal behaviors through full-flow traversal.

In addition, the historical business behavior data of the target merchant can also comprise external evaluation information, and based on the external evaluation information, the safe and reliable score evaluation can be carried out more objectively. For example, the evaluation condition of the consumer to the target merchant can be obtained through the online public opinion, the number of negative evaluations is extracted and counted from the evaluation condition, and the safe and reliable score evaluation model evaluates according to the number of negative evaluations and the content quality; also for example, complaints received by the subject business may be obtained from the regulatory authorities, and the risk that the subject business may have in the business process may be objectively assessed from the consumer perspective. It should be noted that, the external data acquired from the network or the regulatory department or the like does not exclude the possibility of a malicious blackening target merchant, and thus the external data needs to be selectively analyzed.

The front business behavior data of the target merchant can comprise information such as the situation that the target merchant actively participates in self wind control management, external front evaluation and the like. For example, if the target merchant has awareness of risk prevention, in order to prevent risks in time, uploading self-business qualification materials, real business condition pictures or videos through the product to estimate risks in advance, and then the safe and reliable score evaluation model gives higher safe and reliable scores based on the front business behavior data. In addition, objective front management behavior data can be obtained from feedback conditions of external groups such as network public opinion and supervision departments.

In some embodiments, the security trust evaluation model is pre-trained in the following manner, including:

constructing training samples, wherein the training samples comprise positive samples and negative samples, the positive samples comprise different types of positive business behavior data, and the negative samples comprise different types of negative business behavior data;

based on the types of the positive operation behavior data and the negative operation behavior data, setting corresponding safe and reliable scores as labels;

training a pre-constructed classification model by using a training sample and a label until a safe and reliable score evaluation model meeting preset conditions is obtained.

Based on the different types of the positive operation behavior data and the negative operation behavior data and the frequency of occurrence of various types of matters, the safe and reliable score labels are reasonably set. After the training sample is input into a pre-constructed classification model, a safe and reliable score evaluation result is obtained, the difference between the safe and reliable score evaluation result and a safe and reliable score label is calculated, and the difference between the safe and reliable score evaluation result and the safe and reliable score label is minimized to be a target training safe and reliable score evaluation model.

In some more specific embodiments, the security trust assessment model may be built based on the structure of the CNN or RNN network.

The risk level of the target merchant can be divided by determining the current safe and reliable score of the target merchant. Therefore, different wind control services are arranged for target merchants based on the risk, and finally, the risk can be reduced or eliminated, so that the benefit loss of the user is avoided.

S104: based on the security trust score, a low-risk merchant and a high-risk merchant are determined.

In some embodiments, prior to determining the low-risk merchant and the high-risk merchant based on the security trust score, further comprising:

Determining a risk-free merchant based on the industry category;

and eliminating the risk-free merchant from the target merchant.

In other words, before determining the risk type of the target merchant, a merchant white list with ensured business authenticity and security is established; merchants in the white list comprise, but are not limited to, public organizations such as schools and hospitals, KA merchants (key merchants) with great advantages of operation scale and flow in platform cooperation merchants, and some special industry qualification and the like, and the white list merchants are low in risk probability or high in risk resistance, do not need wind control service to a great extent, and can be classified by confirming the industry category of the target merchant, so that forced wind control is avoided, and user experience and working efficiency are optimized.

The basic information of the target merchant can be obtained by basic information data provided when the target merchant signs up with the platform, so as to be used as the authenticity guarantee to be checked and fused with the business behavior data of the target merchant, and the real business scene of the target merchant is restored to identify hidden risks. The basic information when the target merchant signs up with the platform includes, but is not limited to, operation environment information, sign-up product content, and business merchant category codes (MCC codes), etc., the MCC codes are used for identifying the transaction environment and the main camping scope and industry category of the target merchant, and are important information for risk management and transaction control of the related transaction facility, and the MCC codes of the target merchant must be consistent with the main camping industry category thereof. The business behavior data of the target merchant includes, but is not limited to, the collection behavior, the business category, and the buyer feedback of the target merchant, and since the content of the business behavior data is described in the above step S100, the relevant parts are referred to each other, and will not be described herein.

The unbiased and accurate business reality major guarantee white list is characterized by the multi-element and objective target business data, the white list business can be accurately identified, the air-out control range is removed, the unnecessary resource loss caused by the fact that a platform is strongly limited for controlling is avoided when the air-out control decision is made, and business operation experience is optimized.

Based on the safe trusted score, a safe trusted score threshold may be preset for dividing low-risk merchants from high-risk merchants. For example, the safe and reliable score threshold may be set to 60 points, if the safe and reliable score of the target merchant is lower than 60 points, the target merchant is determined to be a high-risk merchant, that is, a serious risk behavior has occurred, and a risk management and control policy needs to be implemented; the merchants with the safe and credible division of 60-80 points are determined to be low-risk merchants, which represent target merchants with risk trends or smaller risks, and some preventive strategies can be adopted to avoid the risks; merchants with a score higher than 80 are identified as current risk-free merchants, and can conduct business normally.

S106: based on the risk category of the low-risk merchant, initiating a preset autonomous risk management interaction service to the low-risk merchant; and updating the safe and trusted score of the low-risk merchant after the low-risk merchant completes the autonomous risk management interactive service.

The autonomous risk management interactive service mainly solves the problems of education prompts and front verification of low-risk merchants with risk trends. Based on the safe and reliable score of the target merchant, a merchant risk credit system is reconstructed by big data, meanwhile, the score value is informed to the authorized target merchant, and risk management interactive services, such as high-risk behavior safety education, interactive video acquisition, legal verification and the like, are pushed to the merchant with low safe and reliable score and no serious risk. The target merchant actively receives security education aiming at specific risk categories, or interacts with the wind control platform by combining measures such as uploading operation video verification, legal person information verification and the like, and automatically improves security credibility to manage risks until the security credibility score is improved to a healthy value, so that the wind control punishment possibility corresponding to the target merchant is reduced, and normal operation can be carried out with confidence; otherwise, if the target merchant does not autonomously manage the risk through the interactive service, the safe and reliable score is maintained at a lower value, and the possibility that the target merchant is subjected to wind control punishment is improved.

By establishing the safe and reliable score of the target merchant, the awareness of maintaining the safe and reliable score of the target merchant can be cultured, so that the target merchant perceives the risk behavior and actively takes measures to avoid the risk behavior. The security awareness of the target merchant is improved, the wind control platform continuously monitors the risk condition of the target merchant, the target merchant is prompted to autonomously participate in the risk management of the target merchant, and the target merchant has initiative in the risk management process, rather than passively receiving platform management and treatment after capturing the risk behaviors.

In some embodiments, after the low-risk merchant completes the autonomous risk management interaction service, updating the security trust score of the low-risk merchant specifically includes:

acquiring business behavior data of a low-risk merchant in the process of completing autonomous risk management interactive service;

inputting the operation behavior data into a safe and reliable score evaluation model;

As described above, the business behavior data of the low-risk merchant in the process of completing the autonomous risk management interactive service can include front business behavior data such as high-risk behavior security education, interactive video acquisition, legal verification and the like, and the higher the completion degree of the target merchant, the higher the evaluation result of the security credibility evaluation model aiming at different types of front business behaviors. For example, the target merchant can perform high-risk behavior security education in a mode of combining video learning and test questions, and the more the video watching time length is or the higher the test question score is, the higher the obtained security and credibility score is; for another example, the target merchant performs risk verification by uploading the camping video, however, if the quality problems such as long too short time or blurred pictures of the video collected by the target merchant cause difficulty in identifying the real operation condition, the effect on updating the safe and reliable score is limited.

The security trust score is updated through the autonomous risk management interactive service of the target merchant, so that the risk and the security of the target merchant are visible, the change of the risk and the security is transparent, the passivity of the target merchant before the risk management control is improved, and the target merchant has sufficient psychological preparation and response time to face the risk management control.

S108: and determining a management and control strategy for the high-risk merchant through a strategy recommendation model based on the historical operation behavior data of the high-risk merchant.

The strategy recommendation model is used for identifying and analyzing the existing risk category according to the collected business data of the high-risk target merchant, so that a proper risk management and control strategy is decided; the input of the management and control strategy is the historical management and control behavior data (including but not limited to data of collection behavior, management type, buyer feedback and the like) of the high-risk target merchant, and the corresponding management and control strategy is output.

In some embodiments, the policy recommendation model is pre-trained in the following manner:

s200: a set of management and control policies is determined based on the historical management and control data.

The history management and control data can comprise characteristics of high-risk merchants subjected to risk management and control and management and control methods decided and implemented by the wind control platform aiming at the high-risk merchants under different conditions, the history management and control data can be obtained by searching a server where the wind control platform is located, the data can be retrieved from other related servers, and the related management and control data can be searched from a cloud server.

The control strategy consists of specific control actions and action periods, and different control strategies can be embodied as different combinations of control actions and action periods; wherein the administration actions include, but are not limited to, limiting certain functions of the corresponding merchant, such as a checkout function, a store up new function, etc. The control strategy can be implemented by the same control action in a non-passing action period, such as limiting the collection function of the corresponding merchant for three days or one week; or different control actions can be implemented in the same period; of course, different control actions can be implemented in different action cycles.

S202: for each management policy in the management policy set, collecting first operational behavior data of the corresponding merchant before the management policy is implemented, and collecting second operational behavior data of the corresponding merchant after the management policy is implemented and complaint data of the corresponding merchant for the management policy.

The first operation behavior data of the corresponding merchant reflect the operation behavior with risk before management and control, and the second operation behavior data can directly reflect the risk situation of the corresponding merchant after the management and control strategy is implemented, and can be specifically represented as the probability of the risk of the corresponding merchant after the management and control, so that the management and control effect of the corresponding management and control strategy is represented. The complaint data comprise the evaluation and feedback of the management and control strategy by the corresponding target merchant, and especially the complaint proposed by the irrational place of the management and control strategy by the corresponding merchant; even the wrong management and control strategy is implemented due to misjudgment of the risk category, so that normal operation and wind control experience of the merchant is affected, and the corresponding merchant is caused to complain.

S204: and determining a control effect score of the control strategy based on the second business behavior data.

The second business behavior data includes, but is not limited to, data of the collection behavior, business category, buyer feedback, etc. of the corresponding merchant after implementing the management and control policy. In some specific embodiments, the feature extraction may be performed on the second business behavior data by a general method, and the feature extracted from the first business behavior data by the same method may be compared, so that the control effect of the control policy is reflected by the comparison result, and the control effect score is determined. The higher the management effect score, the better the management effect of the management strategy is indicated.

S206: based on the second operational performance data, environmental information is determined.

The environment information is indispensable data in the reinforcement learning process and is used for reflecting the management and control effect of the management and control strategy so as to promote updating and learning of the strategy recommendation model and generate a more scientific and effective management and control strategy.

S208: based on the management effect score and the complaint data, an incentive signal is determined.

In some embodiments, the excitation signal specifically includes a reward signal and a penalty signal.

Because the management and control effect score can reflect the effect of the management and control strategy to a certain extent, when the management and control effect score is higher, a reward signal can be determined so as to mark effective management and control strategy parameters for learning; when the score of the management and control effect is lower, the management and control effect of the management and control strategy is limited, the management and control strategy does not need to be learned in the subsequent model learning process, and particularly when the management and control strategy carries out error management and control on corresponding merchants due to misjudgment risks, the score of the management and control effect is correspondingly lower when the merchants complain, therefore, punishment signals can be determined, and the strategy recommendation model is reminded to reduce the occurrence of the errors.

The determination of the excitation signal is helpful for referencing and learning historical data in the process of updating the strategy recommendation model, and a more effective and targeted management and control strategy is generated, so that the training efficiency of the strategy recommendation model is improved, and the wind control experience of a merchant is optimized.

S210: and performing reinforcement learning based on the environmental information and the excitation signal until a strategy recommendation model meeting preset conditions is obtained.

Reinforcement learning is a mechanism by which an agent maximizes returns or achieves a specific goal through learning strategies in the course of interaction with an environment; the agent may be hardware, such as a machine, or may be software, such as a computer program. In the embodiment of the specification, the policy recommendation model may be a classification model, and for each round of training process, the policy recommendation model predicts corresponding control effects after implementing different control policies on the corresponding high-risk merchant characteristics, where the control effects may be estimated based on environmental information and excitation signals; the management and control strategies with the best effect are selected through comparison of the management and control effects to learn, and a new management and control strategy is generated; repeating the steps to update the strategy recommendation model for a plurality of rounds, and enhancing the rationality of the generated management and control strategy in one-time learning until the strategy recommendation model meeting the preset conditions is obtained, thereby completing the training of the strategy recommendation model.

In some embodiments, the policy recommendation model satisfying the preset condition may be obtained by presetting a score threshold for the management and control effect score, and if the management and control effect score obtained after the management and control policies generated by the policy recommendation model for different merchants are implemented can be stably higher than the score threshold, training of the policy recommendation model may be ended, and a policy recommendation model capable of making a scientific decision may be obtained.

In some more specific embodiments, the policy recommendation model may be built based on an MLP network structure, and the reinforcement learning method employed may be a PPO reinforcement learning method.

And recommending at least one corresponding management and control strategy according to the historical operation behavior data of the high-risk merchant through a pre-trained strategy recommendation model, respectively predicting management and control effects and user experience effects after implementing the management and control strategy, evaluating each management and control strategy in a quantized form, and selecting the management and control strategy which is most in line with the management and control expectation so as to determine the implementation of the high-risk merchant. The strategy recommendation model considers the risk control result and the feedback of the user, and can achieve the balance between wind control and user experience better.

In some embodiments, the abtest method may be adopted to make decisions of the management and control policies, that is, for different management and control policies, the actual management and control effects and the user experience effects of each management and control policy are respectively evaluated through random traffic respectively, and comparison is performed, so that more effective management and control policies are scientifically decided in a practical manner.

In some embodiments, after implementing the management and control policy on the high-risk merchant, the management and control prompt information is fed back to the high-risk merchant, and a complaint channel entrance link is embedded in a display page of the management and control prompt information.

In some more specific embodiments, the regulatory prompt is used to describe a business risk behavior category for the high-risk merchant; and displaying the complaint channel entry links embedded in the pages and used for pointing to the complaint channels corresponding to the business risk behavior categories.

When the control prompt information is fed back to the high-risk merchant, real-time service link guiding can be realized. Firstly, data are exchanged through a system where the ventilation control platform is located and other related systems, so that merchants can dynamically interact with the ventilation control platform in real time after being controlled, and specific information such as the controlled function, the control reason and the like can be obtained. And then, embedding the complaint channel entrance links into the display page of the management and control prompt information, so that the real-time guidance of the wind control service is realized, the difficulty of finding the complaint channel by merchants and the jump and break points on the complaint path are reduced, and the user experience is optimized while the complaint efficiency is improved.

Taking a merchant collection scene as an example, in payment scenes such as code scanning payment, transfer, red package, online shopping and the like, the wind control platform limits the collection function of a high-risk target merchant, not only transmits payment interception error reporting information to a user terminal of a payer, but also transmits the payment interception error reporting information to the payee, thereby opening up a risk platform and a payment system, enabling the target merchant serving as the payee to receive a risk management and control prompt in real time so as to enter a complaint link and an audit link to autonomously remove risks.

In some embodiments, the wind control platform directs the high risk target merchant to the security service battle site for complaints. In the security service array, the treatment record of the target merchant is displayed in a structured way, and the management and control information including the platform limit type, the management and control reason, the management and control period and the like can reduce the understanding cost of the managed and controlled target merchant and realize the transparency of the wind control link. In addition, for target merchants who have submitted complaint requests and enter the complaint flow, the security service array displays information such as complaint problem processing timeliness, audit data overrule reasons, requirements and the like on a specific page, so that the flow property and the complaint passing rate of the complaint flow are enhanced.

In the complaint process, the target merchant needs to submit complaint materials such as business license, shop signboards, legal person identity information and the like which can prove the legal operation of the target merchant according to requirements, the materials comprise but are not limited to texts, pictures, files and the like, and the materials enter an auditing system for auditing after being uploaded, and an artificial auditing mode can be adopted, and intelligent auditing and artificial auditing can be combined.

In some embodiments, a self-help complaint channel of the merchant is further provided, so that special merchants such as merchants with low risk concentration or mixed business merchants are considered, and various dynamic interactive complaint means such as interactive intelligent video verification business scenes, video commitment standard business and the like are expanded to realize complaint schemes based on different characteristics of the merchant.

In some embodiments, referring to fig. 5, after complaining, the target merchant further includes:

s300: acquiring complaint materials submitted by high-risk merchants through a complaint channel;

s302: pre-auditing the materials meeting the pre-auditing conditions in the application materials based on a preset auditing model;

s304: if the pre-verification result is not passed, the non-passed verification content is fed back to the high-risk merchant for modification;

s306: if the pre-checking result is passed, the complaint material is transferred to a manual checking node for continuous checking.

The complaint material submitted by the high-risk merchant should include at least one complaint picture, such as business license, shop signer or legal identity card, and the like, which can prove the operational qualification compliance and the authenticity of the operational activity, so that the auditing system can audit the authenticity, the integrity and the validity of the merchant qualification according to the supervision requirement and the wind control requirement, including but not limited to whether the merchant qualification is counterfeit or falsified by other people, and whether the relevant certificate, such as the business license, is in the usable term or not, and whether the qualification material provided by the merchant is complete or not.

Optionally, the pre-audit process includes, but is not limited to, a sub-process of picture quality pre-audit, picture type pre-audit, picture text pre-audit, icon pre-audit, and the like, as well as any suitable combination of the above sub-processes. The quality evaluation is carried out on the submitted pictures through the picture quality pre-examination, so that the picture quality problems such as incomplete, overexposure, blurring and the like can be detected, the problems are fed back to high-risk merchants, and the high-risk merchants are guided to reload the pictures with the mass transfer quantity compliance. The picture type pre-checking checks whether the picture is uploaded at the correct position or not for the high-risk merchant, and when a plurality of pictures of different types are required to be uploaded in some complaint schemes, if the picture type is uploaded incorrectly, the checking is not passed, for example, a legal identity card picture is submitted at a business license uploading position. The picture text pre-checking obtains text content information in a picture uploaded by a high-risk merchant through OCR text recognition, and takes business license as an example, information such as a merchant name, a merchant type, a merchant address, a legal person, a business deadline, an operating range, a registration unit and the like in the picture text pre-checking can be identified, and whether the business deadline is expired, whether the business license is logged off or is suspended, whether the legal person is consistent or not can be determined by checking with the merchant information recorded in a database of a third party authority which issues and manages the business license, so that whether the pre-checking passes or not is determined. And the icon pre-checking judges whether the picture is consistent with a preset template through icon acquisition, feature extraction and feature matching, for example, the positions and the sizes of the official stamps in business licenses of the same type of merchants are similar.

When the pre-audit is not passed, the audit system can reject the reasons and the requirements are displayed to the high-risk merchant in detail so as to guide the high-risk merchant to re-submit the complaint materials after modification, and the audit passing rate is improved. If the condition that the multiple pre-audits are not passed occurs, the audit system directly arranges a manual audit link.

The method can cover audit points which cannot be supported by manual audit through the combination of intelligent pre-audit and manual audit, real-time feedback of risks existing in complaint materials is achieved, managed and controlled high-risk merchants are helped to recover operation and development faster, good user experience is guaranteed while overall audit efficiency is improved, on the other hand, manual audit cost can be further solved, cost is reduced, and efficiency is improved.

In some embodiments, after executing the control strategy for the high-risk merchant, acquiring the control behavior data of the high-risk merchant for the control strategy through a node preset in the control channel; determining a solution behavior path of the high-risk merchant based on the node where the solution behavior data occurs; and determining user experience barrier points on the solution control behavior path based on the solution control behavior path and a preset solution control path.

The nodes preset in the control channel are used for collecting data related to the high-risk merchant in the process of implementing the control strategy, wherein the data comprise multi-mode and multi-source control behavior data such as risk data, service guide data, merchant behavior data, complaint data and the like. By burying the nodes in each link of the solution control channel, multiple data can be fused to restore the solution control action path of the high-risk merchant, and a large line graph of the path of the merchant is built, so that the execution condition and the obstacle point condition of each link of the merchant on the solution control full link, such as complaints, audits and the like, can be clearly displayed, and compared with a preset ideal path which can be successfully subjected to solution control, whether the solution control action of the high-risk merchant is reasonable or not can be detected, and the specific link corresponding to the obstacle point which obstructs the user experience can be obtained, so that the medicine can be applied to the symptoms. The wind control system can be further optimized aiming at the detected user experience obstacle points, so that wind control experience feedback data operation of merchants is realized.

According to the wind control service providing method provided by one or more embodiments of the specification, the risk condition of the target merchant is determined by evaluating the security and reliability points, education prompts and front-end problem solving are carried out on the target merchant with the existing risk trend, the target merchant is promoted to actively participate in self risk management, and risk behaviors are avoided; the method has the advantages that scientific management and control are carried out on target merchants with risk behaviors, management and control prompt information is directly and effectively transmitted to the risk merchants, and complaint channels and pre-examination links are provided, so that the understandability of merchant management and control information and the fluency of complaint links are improved, different complaint schemes are recommended based on merchant values and management and control means, and the efficiency of post-air control service is improved. In addition, after the management and control strategy is executed for the high-risk merchant, the solution control behavior path is obtained and analyzed, so that the wind control experience barrier point of the target merchant can be restored, and evaluation feedback data operation aiming at the management and control strategy is realized.

In another embodiment of the present specification, a wind control service providing apparatus is provided. Fig. 6 is a block diagram schematically illustrating a structure of an air control service providing apparatus according to one or more embodiments of the present disclosure.

As shown in fig. 6, includes:

a data acquisition unit 40 configured to acquire historical operation behavior data of a target merchant within a preset period of time;

a safe trusted score module 42 configured to input the historical operational behavior data into a pre-trained safe trusted score evaluation model, and determine a current safe trusted score of the target merchant based on an evaluation result of the safe trusted score evaluation model; the method comprises the steps that a target merchant receives a preset autonomous risk management interaction service, and the safety and credibility of the target merchant are updated in response to the completion of the preset autonomous risk management interaction service by the target merchant;

a wind control policy generation module 44 configured to determine low risk users and high risk users based on the security trust score; initiating a preset autonomous risk management interaction task based on the risk category of the low-risk merchant; based on historical operation behavior data of high-risk merchants, determining a management and control strategy for the high-risk merchants through a strategy recommendation model, and initiating high-risk wind control tasks;

a wind control engine 46 configured to provide preset autonomous risk management interaction services to low risk merchants in response to the autonomous risk management interaction tasks; and in response to the high risk wind control task, enforcing a management policy for the high risk user.

In order for the data acquisition unit to acquire historical business behavior data of a more representative target merchant, so that the historical business behavior data can describe as many merchant business conditions as possible, the interval of the preset time period is not too short, and meanwhile, the storage cost and the operation rate should be considered, and a proper preset time period is selected. The historical business behavior data of the target merchant can reflect information such as the collection behavior, business category, buyer feedback and the like of the merchant, is favorable for restoring the real business scene of the target merchant, and can evaluate the risk and credibility of the target merchant more accurately based on multi-element and objective data. The data reflecting the collecting behavior of the target merchant includes, but is not limited to, collecting product information, collecting mode, single collecting amount, collecting frequency, daily business amount, etc., and the data obtaining unit may obtain the data related to the collecting behavior from the server corresponding to the collecting terminal. In addition, the historical business behavior data of the target merchant can also comprise buyer LBS aggregation conditions, and can reflect the main business place and audience geographic position distribution; the abnormal operation sequence of the merchant can record abnormal behaviors such as abnormal large-amount transactions, unqualified inspection quality and the like through monitoring the daily business behaviors of the merchant.

The historical business behavior data of the target merchant can be divided into positive business behavior data and negative business behavior data, and the information which can be obtained by analyzing the negative business behavior data of the merchant includes, but is not limited to, merchant business conditions, abnormal behaviors and external evaluations. The merchant operation status information may include whether a transaction dispute is generated, such as whether a collection amount is inconsistent with a contract, whether repeated collection is performed, whether goods or services are seriously inconsistent with descriptions, so that consumer complaints are caused, and the like, and the safe and trusted score module evaluates safe and trusted scores according to the severity of the dispute so as to identify the condition that the black ash merchant obtains illegal benefits by doing hands and feet in a transaction mode; the method can also comprise the steps of judging whether the commodity operated by the target merchant belongs to a forbidden commodity or a limited commodity, judging whether the limited commodity is sold according to relevant regulations, judging whether the target merchant has false marketing phenomenon in the propaganda process, judging whether the target merchant has dishonest conditions such as non-regulated performance after signing with the platform; for another example, the bottom line risk development and investigation of the target merchant are performed, and related consumer complaint conditions are obtained, and the safe and reliable sub-module estimates the risk bearing capacity of the target merchant according to the bottom line risk development and investigation, so that the safe and reliable sub-module evaluates the safe and reliable sub-module of the target merchant.

On the other hand, the safe and trusted score module can also evaluate safe and trusted scores according to abnormal behaviors of the merchant, such as detecting whether abnormal transactions (such as large-amount transactions) exist in a merchant transaction scene and the frequency and the amount of the abnormal transactions; the quality and the checking result of the qualification materials submitted by the target merchant during annual inspection and/or patrol inspection can influence the checking result by the authenticity and the validity of the qualification materials; and then, capturing the behavior sequence of the target merchant through a related algorithm, for example, capturing the behavior of the target merchant in the links of goods incoming, processing, propaganda, sales and the like, and further, recognizing abnormal behaviors through full-flow traversal.

The front business behavior data of the target merchant can comprise information such as the situation that the target merchant actively participates in self wind control management, external front evaluation and the like. For example, if the target merchant has awareness of risk prevention, in order to prevent risks in time, uploading self-business qualification materials, real business condition pictures or videos through the product to estimate risks in advance, and then the safe and reliable score module gives higher safe and reliable scores based on the front business behavior data. In addition, objective front management behavior data can be obtained from feedback conditions of external groups such as network public opinion and supervision departments.

In some embodiments, the safe trusted score evaluation model in the safe trusted score module is pre-trained in the following manner, including:

The safe and reliable score module can divide the risk of the target merchant by determining the current safe and reliable score of the target merchant. Therefore, different wind control services are arranged for target merchants based on the risk, and finally, the risk can be reduced or eliminated, so that the benefit loss of the user is avoided.

The autonomous risk management interactive service mainly solves the problems of education prompts and front verification of low-risk merchants with risk trends. The security and credibility sub-module reconstructs a merchant risk credit system by using big data, informs an authorized target merchant of the score value, and the wind control strategy generation module pushes risk management interactive service for merchants with low security and credibility scores and no serious risk, and executes the risk management interactive service through a wind control engine, such as high-risk behavior security education, interactive video acquisition, legal personal verification and the like. The target merchant actively receives security education aiming at specific risk categories, or interacts with the wind control engine by matching with measures such as uploading operation video verification, legal person information verification and the like, and automatically improves security credibility to manage risks until the security credibility value is improved to a healthy value, so that the wind control punishment possibility corresponding to the target merchant is reduced, and normal operation can be carried out with confidence; otherwise, if the target merchant does not autonomously manage the risk through the interactive service, the safe and reliable score is maintained at a lower value, and the possibility that the target merchant is subjected to wind control punishment is improved.

By establishing the safe and reliable score of the target merchant, the awareness of maintaining the safe and reliable score of the target merchant can be cultured, so that the target merchant perceives the risk behavior and actively takes measures to avoid the risk behavior. The security awareness of the target merchant is improved, the wind control engine continuously monitors the risk condition of the target merchant, and the target merchant is prompted to autonomously participate in own risk management, so that the target merchant has initiative in the risk management process, and is not passively subjected to management and treatment after capturing the risk behaviors.

In some embodiments, the secure trusted partition module specifically includes:

As described above, the business behavior data of the low-risk merchant in the process of completing the autonomous risk management interactive service can include front business behavior data such as high-risk behavior security education, interactive video acquisition, legal verification and the like, and the higher the completion degree of the target merchant, the higher the evaluation result of the security trusted partition module is for different types of front business behaviors.

In some embodiments, the wind control policy generation module, prior to determining the low-risk merchant and the high-risk merchant based on the security trust score, further comprises:

the wind control strategy generation module determines the industry category of the target merchant based on the basic information and the business behavior data of the target merchant;

determining a risk-free merchant based on the industry category;

and eliminating the risk-free merchant from the target merchant.

Based on the safe trusted score, the wind control strategy generation module can preset a safe trusted score threshold value for dividing low-risk merchants and high-risk merchants. For example, the safe and reliable score threshold may be set to 60 points, if the safe and reliable score of the target merchant is lower than 60 points, the target merchant is determined to be a high-risk merchant, that is, a serious risk behavior has occurred, and a risk management and control policy needs to be implemented; the merchants with the safe and credible division of 60-80 points are determined to be low-risk merchants, which represent target merchants with risk trends or smaller risks, and some preventive strategies can be adopted to avoid the risks; merchants with a score higher than 80 are identified as current risk-free merchants, and can conduct business normally.

The wind control engine generates a corresponding management and control strategy through a strategy recommendation model deployed in the wind control engine, and the strategy recommendation model is used for identifying and analyzing the existing risk category according to the collected management data of the high-risk target merchant so as to determine a proper management and control strategy; the input of the management and control strategy is the historical management and control behavior data (including but not limited to data of collection behavior, management type, buyer feedback and the like) of the high-risk target merchant, and the corresponding management and control strategy is output.

In some embodiments, the policy recommendation model in the wind control engine is pre-trained in the following manner:

S200: determining a management and control strategy set based on the history management and control data;

s202: for each management policy in the management policy set, collecting first operation behavior data of a corresponding merchant before the management policy is implemented, and collecting second operation behavior data of the corresponding merchant after the management policy is implemented and complaint data of the corresponding merchant on the management policy;

s204: determining a management and control effect score of the management and control strategy based on the second business behavior data;

s206: determining environmental information based on the second business activity data;

s208: determining an excitation signal based on the management and control effect score and the complaint data;

The wind control engine recommends at least one corresponding control strategy according to historical operation behavior data of the high-risk merchant, predicts the control effect and the user experience effect after the control strategy is implemented respectively, evaluates each control strategy in a quantized mode, and accordingly selects the control strategy which is most in line with control expectations to determine to be implemented for the high-risk merchant, the risk control result is considered, feedback of a user is considered, and finally balance between wind control and user experience can be well achieved.

In some embodiments, the wind control engine may adopt the abtest method to make decisions of the control policies, that is, for different control policies, the actual control effect and the user experience effect of each control policy are respectively evaluated by implementing random traffic, and comparison is made, so that a more effective control policy is scientifically decided in a practical manner.

In some embodiments, after implementing the management and control policy on the high-risk merchant, the wind control engine feeds back management and control prompt information to the high-risk merchant, and embeds a complaint channel entrance link in a display page of the management and control prompt information.

When the wind control engine feeds back the control prompt information to the high-risk merchant, real-time service link guidance can be realized. Firstly, data are exchanged through a system where the ventilation control engine is located and other related systems, so that merchants can dynamically interact with risk information in real time after being controlled by the ventilation control engine, and specific information such as the controlled function, the control reason and the like can be obtained. And then, embedding the complaint channel entrance links into the display page of the management and control prompt information, so that the real-time guidance of the wind control service is realized, the difficulty of finding the complaint channel by merchants and the jump and break points on the complaint path are reduced, and the user experience is optimized while the complaint efficiency is improved.

In some embodiments, the wind control engine directs the high risk target merchant to the security service battle site for complaints. In the security service array, the treatment record of the target merchant is displayed in a structured way, and the management and control information including the platform limit type, the management and control reason, the management and control period and the like can reduce the understanding cost of the managed and controlled target merchant and realize the transparency of the wind control link. In addition, for target merchants who have submitted complaints and enter the complaint flow, the wind control engine displays information such as complaint problem processing timeliness, data checking and rejecting reasons, requirements and the like on a specific page of the security service array, so that the flow property and the complaint passing rate of the complaint flow are enhanced.

In some embodiments, the wind control engine also provides a self-help complaint channel for the commercial tenant, so that special commercial tenants such as commercial tenants with low risk concentration or mixed business commercial tenants are considered, and various dynamic interactive complaint means such as interactive intelligent video verification business scenes, video promise standard business and the like are expanded, so that complaint schemes based on commercial tenant characteristics are realized.

In some embodiments, the wind control engine further comprises:

The complaint material submitted by the high-risk merchant should include at least one complaint picture, such as business license, shop signer or legal identity card, etc., which can prove the operational compliance and authenticity of the operational activity, so that the auditing system of the air control engine can conduct authenticity, integrity and validity auditing on the merchant qualification according to the supervision requirement and the air control requirement, including but not limited to whether the merchant qualification is counterfeit or impostor others, and whether the relevant certificate, such as business license, is in the usable term, and whether the qualification material provided by the merchant is complete.

The wind control engine can cover audit points which cannot be supported by manual audit through a mode of combining intelligent pre-audit and manual audit, real-time feedback of risks existing in complaint materials is achieved, managed and controlled high-risk merchants are helped to recover operation and development faster, good user experience is guaranteed while overall audit efficiency is improved, on the other hand, manual audit cost can be further solved, cost is reduced, and efficiency is improved.

In some embodiments, after the wind control engine executes the control strategy for the high-risk merchant, acquiring the control behavior data of the high-risk merchant aiming at the control strategy through the node preset in the control channel; determining a solution behavior path of the high-risk merchant based on the node where the solution behavior data occurs; and determining user experience barrier points on the solution control behavior path based on the solution control behavior path and a preset solution control path.

The nodes preset in the control channel are used for collecting data related to the high-risk merchant in the process of implementing the control strategy, wherein the data comprise multi-mode and multi-source control behavior data such as risk data, service guide data, merchant behavior data, complaint data and the like. The wind control engine can restore the solution control action path of the high-risk merchant by fusing various data through burying the nodes in each link of the solution control channel, and construct a large line graph of the path of the merchant, so that the execution condition and the obstacle point condition of each link of the merchant on the solution control full link, such as complaints, audits and the like, can be clearly displayed, and are compared with the preset ideal path which can be successfully solved, whether the solution control action of the high-risk merchant is reasonable or not can be detected, and the specific link corresponding to the obstacle point which obstructs the user experience can be obtained, so that the medicine can be issued for symptoms. Aiming at the detected user experience obstacle point, the wind control engine can obtain further optimization, so that wind control experience feedback data operation of the merchant is realized.

One embodiment in the present specification also provides an electronic device, including:

a processor;

a memory for storing processor-executable instructions;

wherein the processor implements the method as described in any one of the wind-controlled service provision methods above by executing executable instructions.

An embodiment in the present specification further provides a computer readable storage medium storing a computer program which when executed by a processor implements a method as set forth in any one of the wind-controlled service providing methods above.

Fig. 7 exemplarily shows a block diagram of an electronic device provided in an embodiment of the present disclosure, which shows a schematic structural diagram of a computer system 500 of a terminal device or a server suitable for implementing an embodiment of the present invention. The terminal device or server shown in fig. 7 is only an example, and should not impose any limitation on the functions and scope of use of the embodiments of the present invention.

In a typical configuration, computer 500 includes one or more processors (CPUs) 502, an input interface 508, an output interface 510, a network interface 506, and a memory 504.

Memory 504 may include non-volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

In particular, according to embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flow chart. The above-described functions defined in the method of the present invention are performed when the computer program is executed by a Central Processing Unit (CPU) 502.

The computer readable medium shown in the present invention may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.

The units or modules involved in the embodiments of the present invention may be implemented in software or in hardware. The described units or modules may also be provided in a processor, for example, as: a processor comprises a data acquisition unit, a safe and reliable sub-module, a wind control strategy generation module and a wind control engine. The names of these units or modules do not constitute limitations on the unit or module itself in some cases, and for example, the data acquisition unit may also be described as "a unit configured to acquire historical business behavior data of a target merchant within a preset period of time".

As another aspect, the present invention also provides a computer-readable medium that may be contained in the apparatus described in the above embodiments; or may be present alone without being fitted into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to include: acquiring historical operation behavior data of a target merchant in a preset time period; determining the current safe and reliable point of the target merchant based on the historical operation behavior data; determining a low-risk merchant and a high-risk merchant based on the security trust score; based on the risk category of the low-risk merchant, initiating a preset autonomous risk management interaction service to the low-risk merchant; after the low-risk merchant completes the autonomous risk management interactive service, updating the safe and reliable score of the low-risk merchant; and determining a management and control strategy for the high-risk merchant through a strategy recommendation model based on the historical operation behavior data of the high-risk merchant. In addition, the strategy recommendation model can be trained directly in the computer readable medium, or can be trained and then loaded into the computer readable medium.

The foregoing describes specific embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can 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 are also possible or may be advantageous. It will also be noted that each block of the figures, and combinations of blocks in the figures, can be implemented by special purpose hardware-based systems which perform the specified functions or operations, or combinations of special purpose hardware and computer instructions.

It should be noted that the above-mentioned embodiments are merely examples of the present invention, and it is obvious that the present invention is not limited to the above-mentioned embodiments, and many similar variations are possible. All modifications attainable or obvious from the present disclosure set forth herein should be deemed to be within the scope of the present disclosure.

Claims

1. A wind control service providing method is applied to a wind control platform and comprises the following steps:

determining a low-risk merchant and a high-risk merchant based on the safe trusted score;

and determining a management and control strategy for the high-risk merchant through a strategy recommendation model based on the historical operation behavior data of the high-risk merchant.

2. The method of claim 1, further comprising, prior to determining the low-risk merchant and the high-risk merchant based on the secure trust score:

determining a risk-free merchant based on the industry category;

and eliminating the risk-free merchant from the target merchant.

3. The method of claim 1, the safe trusted score evaluation model being pre-trained in a manner comprising:

4. The method of claim 1, after the low-risk merchant completes the autonomous risk management interaction service, updating the security trust score of the low-risk merchant, comprising:

5. The method of claim 1, the policy recommendation model is pre-trained in the following manner:

6. The method of claim 1, further comprising:

7. The method of claim 6, wherein the regulatory prompt is used to describe a business risk behavior category of the high-risk merchant; and the complaint channel entrance links embedded in the display page are used for pointing to the complaint channels corresponding to the business risk behavior categories.

8. The method of claim 7, further comprising:

9. The method of claim 1, further comprising:

10. A wind-controlled service providing apparatus comprising:

11. An electronic device, comprising:

a processor;

a memory for storing the processor-executable instructions;

wherein the processor is configured to implement the method of any of claims 1-9 by executing the executable instructions.

12. A computer readable storage medium storing a computer program which, when executed by a processor, implements the method of any one of claims 1-9.