Risk control visualization processing

Abstract

Disclosed in the present disclosure are a risk control visualization processing method and apparatus, a storage medium, and an electronic device. The method comprises: obtaining risk control decision event information, and determining at least one visualization process element corresponding to the risk control decision event information; performing risk control decision generation processing on the basis of each visualization process element to obtain a risk control decision process system; and performing risk control transaction processing on the basis of the risk control decision process system, and displaying a system risk control decision diagram by means of the risk control decision process system.

Description

Risk control visualization processing TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computer, and particularly relates to risk control visualization processing. BACKGROUND

[0002] Risk control (referred to as risk control) engines on domestic platform services are mainly based on risk control rule engines, and functions such as data collection, three-party service integration and algorithm execution are integrated in the risk control decision system. After the development of the risk control decision system, the risk control decision system is equivalent to a black box for non-system developers, and it is difficult for the platform service side to master the strategy running process of the risk control decision system, and it is also difficult to restore the transaction risk control process from the system execution log of the risk control decision system. SUMMARY

[0003] The present disclosure provides a risk control visualization processing method, device, storage medium and electronic equipment, and the technical solution is as follows.

[0004] In a first aspect, the present disclosure provides a risk control visualization processing method, and the method comprises: acquiring risk control decision event information, determining at least one visualization process element corresponding to the risk control decision event information; performing risk control decision generation processing based on each visualization process element to obtain a risk control decision process system; performing risk control transaction processing based on the risk control decision process system and displaying a system risk control decision graph through the risk control decision process system.

[0005] In a second aspect, the present disclosure provides a risk control visualization processing device, and the device comprises: an element determination module configured to acquire risk control decision event information and determine at least one visualization process element corresponding to the risk control decision event information; a decision generation module configured to perform risk control decision generation processing based on each visualization process element to obtain a risk control decision process system; and a transaction processing module configured to perform risk control transaction processing based on the risk control decision process system and display a system risk control decision graph through the risk control decision process system.

[0006] In a third aspect, the present disclosure provides a computer storage medium, and the computer storage medium stores at least one instruction, the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present disclosure.

[0007] In a fourth aspect, the present disclosure provides a computer program product, and the computer program product stores at least one instruction, the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present disclosure.

[0008] In a fifth aspect, the disclosure provides an electronic device, which can include a processor and a memory; wherein the memory stores a computer program, the computer program is adapted to be loaded by the processor and execute the method steps of one or more embodiments of the disclosure.

[0009] The technical solutions provided by some embodiments of the disclosure have at least the following beneficial effects: in one or more embodiments of the disclosure, the service platform obtains risk control decision event information, determines at least one visual process element corresponding to the risk control decision event information, then performs risk control decision generation processing based on each visual process element to obtain a risk control decision process system, and finally performs risk control transaction processing based on the risk control decision process system and displays a system risk control decision graph through the risk control decision process system. For the non-system R&D side of the customer service end, the user end, the supervision party, and the like, the risk control decision process system displays the system risk control decision graph, greatly enhances the risk control explainability, restores the system risk control decision process through the visual means, and improves the intelligent degree of the risk control decision. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the disclosure, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the disclosure, and other drawings can also be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0011] Fig. 1 is a scene schematic diagram of a risk control visualization processing system provided by the disclosure;

[0012] Fig. 2 is a flow schematic diagram of a risk control visualization processing method provided by the disclosure;

[0013] Fig. 3 is a flow schematic diagram of another risk control visualization processing method provided by the disclosure;

[0014] Fig. 4 is a schematic diagram of a system risk control decision graph provided by the disclosure;

[0015] Fig. 5 is a flow schematic diagram of another risk control visualization processing method provided by the disclosure;

[0016] Fig. 6 is a structural schematic diagram of a risk control visualization processing device provided by the disclosure;

[0017] Fig. 7 is a structural schematic diagram of an electronic device provided by the disclosure. DETAILED DESCRIPTION

[0018] The technical solutions in the present disclosure will be clearly and completely described in combination with the drawings in the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present disclosure.

[0019] In the description of the present disclosure, it should be understood that the terms "first", "second" and the like are used only for descriptive purposes, and cannot be construed as indicating or implying relative importance. In the description of the present disclosure, it should be noted that, unless otherwise explicitly specified and limited, "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units not listed, or optionally includes other steps or units inherent to the process, method, product or device. The specific meaning of the above terms in the present disclosure can be understood by those skilled in the art according to the specific circumstances. In addition, in the description of the present disclosure, "multiple" means two or more, unless otherwise specified. The association relationship between the associated objects is described, which means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after it.

[0020] In the related art, since the currently developed risk control decision system is equivalent to a black box, the risk control strategy execution process corresponding to the risk control decision system is a high-density calculation rule based on big data. For non-system development side, such as customer service side, user side, regulatory side, etc., the interpretability is poor, which may cause difficult complaint handling, lack of transparency of supervision, and other phenomena in some cases.

[0021] The present disclosure will be described in detail below in combination with specific embodiments.

[0022] Please refer to FIG. 1, which is a scene schematic diagram of a risk control visualization processing system provided by the present disclosure. As shown in FIG. 1, the risk control visualization processing system can at least include a client cluster and a service platform 100.

[0023] The client cluster can include at least one client, as shown in FIG. 1, specifically including a client 1 corresponding to a user 1, a client 2 corresponding to a user 2,..., and a client n corresponding to a user n, n is an integer greater than 0.

[0024] The clients in the client cluster can be electronic devices with communication functions, including but not limited to wearable devices, handheld devices, personal computers, tablet computers, vehicle-mounted devices, smart phones, computing devices, or other processing devices connected to wireless modems, etc. Electronic devices can be called different names in different networks, such as user equipment, access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent or user device, cellular phone, cordless phone, personal digital assistant (PDA), electronic device in 5G network or future evolution network, etc.

[0025] The service platform 100 can be a separate server device, such as a rack-mounted, blade, tower, or cabinet server device, or a hardware device with strong computing power, such as a workstation, a mainframe computer, etc. The service platform 100 can also be a server cluster composed of multiple servers. The servers in the service cluster can be composed in a symmetrical manner, where each server is functionally and positionally equivalent in the transaction link, and each server can independently provide services externally. The independent service provision can be understood as not requiring the assistance of another server.

[0026] In one or more embodiments of the present disclosure, the service platform 100 can establish a communication connection with at least one client in the client cluster, and complete the interaction of data in the risk control visualization processing process based on the communication connection.

[0027] It should be noted that the service platform 100 and at least one client in the client cluster establish a communication connection through a network for interactive communication, where the network can be a wireless network or a wired network. The wireless network includes but is not limited to a cellular network, a wireless local area network, an infrared network or a Bluetooth network. The wired network includes but is not limited to an Ethernet, a universal serial bus (USB) or a controller area network. In one or more embodiments of the specification, technologies and / or formats including Hyper Text Mark-up Language (HTML), Extensible Markup Language (XML) and the like are used to represent data (such as target compressed packages) exchanged through the network. In addition, all or some links can be encrypted using conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec) and the like. In other embodiments, custom and / or dedicated data communication technologies can be used instead of or in addition to the above data communication technologies.

[0028] The risk control visualization processing system embodiments provided by the present disclosure belong to the same concept as the risk control visualization processing method in one or more embodiments. The execution subject corresponding to the risk control visualization processing method in one or more embodiments of the specification can be the service platform 100 described above, which is determined based on the actual application environment. The implementation process of the risk control visualization processing system embodiment can be seen from the method embodiments described below, which will not be described here.

[0029] Based on the scene diagram shown in FIG. 1, the risk control visualization processing method provided by one or more embodiments of the present disclosure will be described in detail.

[0030] Please refer to FIG. 2, which is a flowchart of a risk control visualization processing method provided by one or more embodiments of the present disclosure. The method can be implemented by relying on a computer program and can run on a risk control visualization processing device based on the von Neumann system. The computer program can be integrated in an application or run as an independent tool application. The risk control visualization processing device can be a service platform.

[0031] Specifically, the risk control visualization processing method includes the following steps.

[0032] S102: Obtain risk control decision event information, and determine at least one visual process element corresponding to the risk control decision event information.

[0033] Risk control decision event information: according to a set of conditions to decide the corresponding risk control process, for example: user portrait is a student, user behavior is credit application, and application product is Huobiao. A set of conditions can correspond to an independent risk control decision event.

[0034] Risk control decision event information is the risk control strategy development information input by the risk control decision maker to the service platform according to the actual risk control demand for developing the risk control decision system.

[0035] Optionally, the risk control decision event information can be in the form of a risk control decision system development document. The risk control decision maker edits the risk control strategy development information for developing the risk control decision system according to the current risk control demand, and inputs the risk control strategy development information to the service platform.

[0036] S104: Based on each of the visual process elements, a risk control decision generation process is performed to obtain a risk control decision process system.

[0038] In one or more embodiments, the corresponding general visual process elements can be set or defined in advance for the risk control strategy decision logic. It can be understood that at least one general visual process element corresponding to the risk control strategy decision logic is set in the system development stage. An element editing interface for changing the risk control strategy decision logic is provided. The at least one general visual process element corresponding to the risk control strategy decision logic is displayed in the element editing interface.

[0039] Optionally, the risk control decision event information can be the description information of the general visual process elements input by the user to build the risk control strategy decision logic. The server can input the risk control decision event information of the general visual process elements by inputting the corresponding human-computer interaction operation to the service platform (the risk control decision event information at this time is the information indicating which general visual process elements are used and how to configure the general visual process elements to build the risk control strategy decision logic), so as to configure the elements of the general visual process elements. The service platform realizes the analysis of at least one visual process element in the risk control decision event information by responding to the human-computer interaction operation (since the risk control decision event information is the description information of the general visual process elements, it is not necessary to analyze the risk control decision semantics).

[0040] Illustratively, the visual process elements can be built into a system risk control decision graph, and then the risk control decision process system is generated after code compilation based on the system risk control decision graph; the (general) visual process elements include but are not limited to sequential flow elements, process node element blocks, and element configuration information corresponding to the process node element blocks.

[0040] The flow node element block includes at least one of a start node element block, a task element block, a gateway element block, an event element block, and an end node element block, and the flow node element block is a degree code system block integrated with a high degree of code function, and the flow node element block is correspondingly configured with a visual system quick pattern on the human-computer interaction interface, and the functions of the start event node element, the task element, the gateway element, the event element, the sequential flow element, and the end event element are encapsulated with corresponding function program codes.

[0041] Optionally, the risk control decision event information can be only a risk control policy decision logic formulated by a service end, and the service platform needs to perform event element analysis processing on the risk control decision event information, analyze risk control decision semantics, and convert the risk control decision semantics into visual flow elements. The visual flow elements can be sequential flow elements indicating a risk control decision flow, flow node element blocks constituting the risk control decision flow, and element configuration information of the flow node element blocks. Subsequently, the visual flow elements can be built into a system risk control decision graph, and then a risk control decision flow system is generated based on the system risk control decision graph after code compilation.

[0042] In a feasible implementation, a risk control decision processing model can be used to perform event element analysis processing on the risk control decision event information to obtain at least one visual flow element. The risk control decision processing model is trained by sample risk control decision event information of a plurality of known visual flow element labels.

[0043] Illustratively, a risk control decision processing model based on a machine learning model can be trained in advance. When the risk control decision event information is a risk control policy decision logic formulated by a service end, the risk control decision event information is input into the risk control decision processing model, the risk control decision semantics are analyzed by the risk control decision processing model, and the risk control decision semantics are converted into visual flow elements, and at least one visual flow element is output.

[0044] In a feasible implementation, a model training process of a risk control decision processing model is exemplified as follows.

[0045] Model creation: An initial risk control decision processing model for a risk control scenario is created based on a machine learning model.

[0046] Sample data acquisition: A large amount of sample data is acquired, and the sample data is composed of sample risk control decision event information input by a sample service end.

[0047] Sample data labeling: based on the demand of the risk control scene, an expert end service is introduced to manually label the sample data with corresponding sample labels. The sample labels are sample visualization process elements, which include sample sequential flow elements, sample process node element blocks, and sample element configuration information corresponding to the sample process node element blocks.

[0048] Model training process: input the sample data into the initial risk control decision processing model for at least one round of model training. The initial risk control decision processing model parses the risk control decision semantic corresponding to the sample risk control decision event information and converts the risk control decision semantic into a predicted visualization process element. Based on the predicted visualization process element and the predicted visualization process element label, a model loss function is used to determine a model loss value. Based on the model loss value, the model parameters of the initial risk control decision processing model are adjusted until the model training end condition is met to obtain a feature matching model.

[0049] Optionally, the model loss function can be a Euclidean distance loss function, a hinge loss function, a cross-entropy loss function, etc.

[0050] Optionally, the model end training condition of the model can include, for example, the value of the loss function being less than or equal to a pre-set loss function threshold, the number of iterations reaching a pre-set number threshold, etc. The specific model end training condition can be determined based on the actual situation, and is not limited here.

[0051] It should be noted that the machine learning model involved in one or more embodiments of the present disclosure includes but is not limited to fitting of one or more of a convolutional neural network (CNN) model, a deep neural network (DNN) model, a recurrent neural network (RNN), an embedding model, a gradient boosting decision tree (GBDT) model, a logistic regression (LR) model, etc.

[0052] S106: based on the risk control decision process system, perform risk control transaction processing and display the system risk control decision graph through the risk control decision process system.

[0053] Specifically, the service platform online the risk control decision process system to the actual risk control transaction environment, and uses the risk control decision process system to process the risk control transaction of the to-be-processed data in the actual risk control transaction environment. In addition, the risk control decision process system is built by using visual process elements, and the visual process elements can intuitively feedback the risk control decision sequence flow direction of the risk control decision process and one or more process node element blocks involved in the risk control decision sequence flow. The risk control decision sequence flow direction and the one or more process node element blocks involved in the risk control decision sequence flow are presented in the system risk control decision graph, and the service platform can display the system risk control decision graph through the risk control decision process system. At this time, the risk control decision process system is no longer a black box for non-developers. Since the system risk control decision graph corresponding to the built risk control decision process system is displayed, the business explainability is high, and the execution result of the risk control strategy can be easily analyzed for non-developers.

[0054] In one or more embodiments of the present disclosure, the service platform obtains risk control decision event information, determines at least one visual process element corresponding to the risk control decision event information, and then generates a risk control decision process system based on each visual process element. The system risk control decision graph is displayed through the risk control decision process system based on the risk control decision process system. The system risk control decision graph is displayed for the non-system development side of the customer service end, the user end, the supervision party, and the like, the risk control explainability is greatly enhanced, the system risk control decision process can be restored through the visual means, and the intelligent degree of the risk control decision is improved.

[0055] Please refer to FIG. 3, which is a flow diagram of another embodiment of a risk control visual processing method according to one or more embodiments of the present disclosure. Specifically, the method includes the following steps.

[0056] S202: Obtain risk control decision event information.

[0057] In one or more embodiments of the present disclosure, the risk control decision processing model can be used to perform event element analysis processing on the risk control decision event information, to obtain at least one visual process element. The risk control decision processing model is obtained by adapting a basic large language generative model to a risk control decision processing scenario. The risk control decision processing model is trained by a plurality of sample risk control decision event information with known visual process element labels.

[0058] S204: Determine a decision event analysis description word for the risk control decision event information.

[0059] The decision event analysis description word is a description word or a prompt word indicating that the risk control decision processing model performs event element analysis processing based on the risk control decision event information.

[0060] S206: input the risk control decision event information and the decision event parsing description word into the risk control decision processing model, perform event element parsing processing on the risk control decision event information through the risk control decision processing model to obtain at least one visual flow element, and output the at least one visual flow element.

[0061] In the present disclosure, the risk control decision processing model is pre-trained, and the risk control decision processing model is obtained after the basic large language model is trained in the risk control decision event element parsing scene migration. The basic large language model can be quickly applied to a new risk control decision scene field without retraining a new model. Only fine-tuning training of the basic large language model is needed. Generally, the basic large language model is adapted to character understanding processing scene. Based on this, the basic large language model (which can be called LLM model) is migrated and converted in the risk control decision event element parsing scene. A multi-modal large-scale language model (which can be called MLLM) compatible with text, image and other multi-modal data can be realized. The multi-modal large-scale language model can perform event element parsing processing. The risk control decision processing model obtained by migrating and training the risk control decision event information as the model input can be called.

[0062] In practical applications, the decision event parsing description word for the risk control decision event information can be determined first. The decision event parsing description word and the risk control decision event information are input into the risk control decision processing model. The risk control decision semantic is parsed and converted into a visual flow element by the risk control decision processing model. At least one visual flow element is obtained by performing event element parsing processing on the risk control decision event information by the risk control decision processing model. The risk control decision processing model outputs at least one visual flow element. The visual flow element includes but is not limited to a sequential flow element, a flow node element block, and element configuration information corresponding to the flow node element block.

[0063] The flow node element block includes but is not limited to at least one of a start node element block, a task element block, a gateway element block, an event element block, and an end node element block. The flow node element block is a high degree of code system block integrated with code functions. The flow node element block is configured with a visual system quick pattern on a human-computer interaction interface. The start event node element function, the task element function, the gateway element function, the event element function, the sequential flow element function, and the end event element function are encapsulated with corresponding function program codes.

[0064] Optionally, the following describes the process of adapting the basic large language generation model to the risk control decision processing scene to obtain the risk control decision processing model.

[0065] A basic large language model is pre-acquired, which can be, for example, a Vicuna-7B model, a ChatGPT large model, a Wenxin Yiyang large model, an Angs large model, etc.

[0066] Then, an event element analysis module for the risk control decision event element analysis scene and a large language generative module corresponding to the basic large language model are determined, and an initial risk control decision processing model including at least the event element analysis module and the large language generative module is created.

[0067] Sample data acquisition: a large amount of sample data is acquired, the sample data is composed of sample risk control decision event information input by a sample server, and a decision event analysis description word (which can be a general decision event analysis description word manually defined) is constructed for the sample data.

[0068] Sample data labeling: based on the demand of the risk control scene, an expert end service is introduced to manually label the sample data with corresponding sample labels by an artificial person, the sample labels are sample visual process elements, and the sample visual process elements include sample sequential flow elements, sample process node element blocks, and sample element configuration information corresponding to the sample process node element blocks.

[0069] Model training process: the sample data is input into the initial risk control decision processing model for at least one round of model training, the initial risk control decision processing model analyzes the risk control decision semantic corresponding to the sample risk control decision event information and converts the risk control decision semantic into predicted visual process elements, determines a model loss value based on the predicted visual process elements and the predicted visual process element labels using a model loss function, adjusts the model parameters of the initial risk control decision processing model based on the model loss value, and stops until the model training end condition is met to obtain a risk control decision processing model.

[0070] Optionally, in the subsequent model training process, a model loss value is determined based on the predicted visual process elements and the predicted visual process element labels using a model loss function, and only the event element analysis module in the initial risk control decision processing model is adjusted in model parameters using the model recognition loss, and the model module parameters of the large language generative module are unchanged.

[0071] S208: Determine the sequential flow elements, the process node element blocks, and the element configuration information corresponding to the process node element blocks from each of the visual process elements, wherein the process node element blocks include a start node element block, a task element block, a gateway element block, an event element block, and an end node element block.

[0072] The flow node element block is correspondingly configured with a visual system quick pattern on the human-computer interaction interface, and the flow node element block such as a start event node element function, a task element function, a gateway element function, an event element function, a sequential flow element function, and an end event element function is pre-encapsulated with a corresponding function program code; for example, as follows.

[0073] Start node element block: indicating the start of the risk control decision process or the starting point in a specific process.

[0074] Task element block: indicating a specific activity or task that needs to be executed in the risk control decision process. In the present disclosure, the task node indicated by the task element block is mainly used to execute code components. Each task node indicated by the task element block represents a system risk control decision behavior, such as querying external data or querying user information.

[0075] Gateway element block: indicating a decision point or branch in the risk control decision process, allowing different flow execution paths to be selected according to specific conditions. In the present disclosure, it is mainly used as a code branching decision or concurrent processing. According to different gateway control types, the network indicated by the gateway element block includes but is not limited to exclusive gateway, parallel gateway, mutual exclusion gateway, etc.

[0076] Event element block: indicating an occurrence or trigger that affects the flow of the risk control decision process, such as a timer event, a message event, or an error event.

[0077] Sequential flow element: indicating the flow of activities and decisions in the process. Based on the sequential flow element, it can be indicated how to connect different elements, i.e., connect the flow node element system block.

[0078] End node element block: indicating the completion or end point of a process or a certain process in the process.

[0079] For example, the element configuration information corresponding to the sequential flow element, the flow node element block, and the flow node element block is determined from the various visual process elements. The element configuration information is used to configure the working parameters of the flow node element block.

[0080] S210: configuring the flow node element block based on the sequential flow element using the element configuration information to obtain a system risk control decision graph.

[0081] In the present disclosure, the sequential flow element indicates how to connect different elements, i.e., connect the flow node element system blocks. Referring to the sequential flow element, the control flow line can be used to connect the various flow node element blocks in the initial system risk control decision graph (usually a blank graph), and the element configuration information is parsed to determine the element configuration parameters of the various flow node element blocks. The element configuration parameters are used to configure the corresponding flow node element blocks. After the above flow system graph configuration process is completed on the initial system risk control decision graph, the system risk control decision graph can be obtained.

[0082] S212: performing system code compilation processing on the system risk control decision graph to generate a risk control decision flow system.

[0083] In the present disclosure, since the flow node element block corresponds to the encapsulation of the corresponding function program code, the function program codes of all flow node element blocks can be combined according to the system risk control decision graph to generate the system program code of the risk control decision flow system. Then, the system program code is compiled to generate the risk control decision flow system.

[0084] S214: obtaining at least one user risk control transaction data in a risk control decision scenario, and calling the risk control decision flow system to perform risk control transaction decision on the user risk control transaction data to obtain risk control system processing information.

[0085] For example, the user risk control transaction data to be processed in the risk control decision scenario is user object, for example, the risk control decision scenario is a credit granting scenario, and the user credit granting transaction data corresponding to the credit granting user object in the credit granting scenario is obtained. The control risk control decision flow system performs credit granting transaction decision based on the user credit granting transaction data to obtain risk control system processing information. The risk control system processing information at least corresponds to the system flow node decision result of each system flow node.

[0086] S216: obtaining the system risk control decision graph corresponding to the risk control decision flow system, determining the system flow node decision result based on the risk control system processing information, adding the system flow node decision result to the system risk control decision graph, and displaying the system risk control decision graph to the server.

[0087] For example, the (default) system risk control decision graph corresponding to the risk control decision flow system is obtained first, and then the system flow node decision result is determined based on the risk control system processing information. The system flow node decision result is added to each system flow node in the (default) system risk control decision graph. The system flow node is a visual system flow node corresponding to each flow node element block. The system risk control decision graph after the system flow node decision result is added is displayed to the server.

[0088] Exemplarily, as shown in FIG. 4, FIG. 4 is a schematic diagram of a system risk control decision diagram, and FIG. 4 shows part of the system risk control decision diagram in the credit system risk control decision diagram in the credit decision scenario. In FIG. 4, after the start node, there are two task nodes corresponding to the two task element blocks: the “check person and review historical information” task and the “person review protection” task. After the two task nodes corresponding to the two task element blocks, there are two decision branches: decision branch 1 has an “exclusive gateway”, and decision branch 2 has a “query user information” task. After the “query user information” task of decision branch 2, there is a parallel event node corresponding to the linking event element block, and the parallel event node is associated with seven parallel task branches: the branch where the “counter-sanction” task is located, the branches where the “intelligence and credit user information” task and the “anti-fraud reporting” task are located, the branch where the “consumption gold new signing feature” task is located, and so on.

[0089] In FIG. 4, there are system flow node decision results such as “hit person review protection rule” system flow node decision result, “hit anti-fraud or counter-sanction” system flow node decision result, and “no external data check” system flow node decision result, which are displayed on the corresponding system flow nodes in the credit system risk control decision diagram.

[0090] Exemplarily, the system risk control decision diagram is displayed for the customer service end, the user end, the regulatory party, and other non-system research and development side risk control decision process systems, and the risk control explainability is greatly enhanced. For example, in the credit granting scenario, the service end associated with the service platform often encounters the following problems when facing the problems of customers: 1. Why is my credit limit 5000? 2. Why was the credit limit reduced? 3. Why was I not granted credit?

[0091] It is difficult for general customer service personnel to accurately locate the risk control decision reason and reply to the customer using the service end. However, in the present disclosure, the system risk control decision diagram has system flow node decision results for each system flow node, and the reasons for the above-mentioned problems are embodied in the flow node decision results. The service end can conveniently locate the decision basis of the risk control decision process system by querying the system risk control decision diagram.

[0092] In one or more embodiments of the present disclosure, the service platform determines at least one visual process element corresponding to the risk control decision event information by acquiring the risk control decision event information, then generates a risk control decision process system based on each visual process element, and then displays a system risk control decision graph based on the risk control decision process system. For the customer service end, the user end, the supervision party and other non-system development side risk control decision process systems, the system risk control decision graph is displayed, the risk control explainability is greatly enhanced, the system risk control decision process can be restored through visual means, and the intelligent degree of risk control decision is improved; and the visual process element can be conveniently determined through the risk control decision processing model, the risk control visual processing process is optimized, and the construction efficiency of the risk control decision process system is greatly improved.

[0093] Please refer to FIG. 5, which is a flow diagram of another embodiment of a risk control visual processing method according to one or more embodiments of the present disclosure. Specifically, the following steps are included.

[0094] S302: In response to the service end initiating a transaction execution query operation for a target user transaction, extracting a target transaction risk control decision snapshot for the target user transaction from the system risk control decision graph based on risk control system processing information.

[0095] Illustratively, such as in the credit granting scenario, the service platform associated with the service end can quickly query the target user transaction corresponding to the user when the service end faces the problem of the customer, such as determining the target user transaction based on the customer information, and then the service end initiates a transaction execution query operation for the target user transaction to the service platform. The transaction execution query operation is used to query the risk control system processing information of the risk control decision process system when processing the target user transaction.

[0096] Generally, it is difficult for customer service personnel using the service end to accurately locate the risk control decision reason and reply to the customer, but in the present disclosure, the risk control visual processing method is used to query the risk control system processing information of the risk control decision process system when processing the target user transaction, and then the target risk control system processing information corresponding to the target user transaction is extracted from the risk control system processing information, and then the target transaction risk control decision snapshot for the target user transaction is extracted from the system risk control decision graph based on the target risk control system processing information.

[0097] The target transaction risk control decision snapshot is a process snapshot of the target risk control decision process triggered by the risk control decision process system when processing the target user transaction, and the target risk control decision process is usually a part of the decision process of the system risk control decision graph corresponding to all risk control decision processes.

[0098] According to the target risk control system processing information, the target risk control decision flow and the target process node decision result of the target process node on the target risk control decision flow are determined from the system risk control decision graph, and then the target transaction risk control decision snapshot is generated based on the target risk control decision flow and the target process node decision result. Each target process node on the target transaction risk control decision snapshot will correspond to the process node decision result of the query user transaction problem cause in the process node decision result, which will be embodied in the process node decision result. The service end can conveniently locate the decision basis of the risk control decision flow system when processing the target user transaction by querying the target transaction risk control decision snapshot.

[0099] In a feasible implementation, the following method can be used.

[0100] C2: The service platform determines the target transaction risk control decision path for the target user transaction based on the risk control system processing information from the system risk control decision graph, and determines the risk control path decision analysis of the target transaction risk control decision path. The target transaction risk control decision path is the process path of the target risk control decision flow triggered by the risk control decision flow system when processing the target user transaction in the complete system risk control decision graph. The service platform can query the risk control system processing information of the risk control decision flow system when processing the target user transaction, and then extract the target risk control system processing information corresponding to the target user transaction from the risk control system processing information. Then, according to the target risk control system processing information, the target transaction risk control decision path triggered by the risk control decision flow system when processing the target user transaction is determined from the complete system risk control decision graph. Then, the risk control path decision analysis of the target transaction risk control decision path is generated. The risk control path decision analysis is an automatic generation of the decision execution analysis description of the target transaction risk control decision path of the risk control decision flow system when processing the target user transaction based on a machine learning model. The risk control path decision analysis can intuitively present the risk control decision logic of the risk control decision flow system when processing the target user transaction, so as to facilitate the service end to intuitively understand the risk control decision logic of the target user transaction.

[0101] Optionally, the system decision explanation model can be pre-trained based on the machine learning model. The system decision explanation model can be used to automatically generate the target transaction risk control decision path and the risk control path decision analysis according to the system risk control decision graph and the risk control system processing information. Specifically, the service platform generates a decision description prompt word for the target user transaction; the decision description prompt word is used to instruct the system decision explanation model to determine the target transaction risk control decision path for the target user transaction from the system risk control decision graph and generate the risk control path decision analysis for the target transaction risk control decision path; the service platform inputs the system risk control decision graph, the risk control system processing information, and the decision description prompt word into the system decision explanation model, determines the target transaction risk control decision path for the target user transaction from the system risk control decision graph based on the risk control system processing information through the system decision explanation model, and determines the risk control path decision analysis based on the target transaction risk control decision path, and outputs the target transaction risk control decision path and the risk control path decision analysis.

[0102] In the present disclosure, the system decision explanation model is pre-trained, and after the system decision explanation model is trained in the risk control decision explanation scene, the basic large language model can be quickly applied to a new risk control decision explanation scene without retraining a new model. Only fine-tuning training of the basic large language model is needed. Generally, the basic large language model is adapted to character understanding processing scene. Based on this, the basic large language model (which can be called LLM model) is migrated and converted in the risk control decision explanation scene to realize a multi-modal large-scale language model (which can be called MLLM) compatible with text, image, and other multi-modal message scenes. The multi-modal large-scale language model can perform risk control path decision analysis. The system risk control decision graph and the risk control system processing information are used as model inputs. The migrated and trained multi-modal large-scale language model can be called a system decision explanation model.

[0103] In practical applications, the decision description prompt word for the target user transaction can be determined first, and the system risk control decision graph, the risk control system processing information, and the decision description prompt word are input into the system decision explanation model. The target transaction risk control decision path for the target user transaction is determined from the system risk control decision graph based on the risk control system processing information, and the risk control path decision analysis is determined based on the target transaction risk control decision path. The target transaction risk control decision path and the risk control path decision analysis are output.

[0104] Optionally, the following explains the process of adapting the basic large language generation model to the risk control decision explanation scene to obtain the system decision explanation model.

[0105] Pre-acquire a basic large language model, such as Vicuna-7B model, ChatGPT large model, Wenxin Yiyang large model, and so on.

[0106] Then, based on the machine learning model, a risk control decision explanation module for the risk control decision explanation scene and a large language generation module corresponding to the basic large language model are created, and an initial system decision explanation model including at least the risk control decision explanation module and the large language generation module is created.

[0107] Sample data acquisition: a large amount of sample data is acquired, the sample data is composed of a sample system risk control decision graph and sample risk control system processing information, and a decision description prompt word for a sample user transaction is constructed for the sample data (which can be a manually customized description word for a target user transaction).

[0108] Sample data labeling: based on the demand of the risk control scene, an expert end service is introduced to manually label the sample data with corresponding sample labels by artificial labeling, the sample labels being transaction risk control decision path labels and risk control path decision analysis labels.

[0109] Model training process: the sample data is input into the initial system decision explanation model for at least one round of model training, the initial system decision explanation model determines a sample transaction risk control decision path for a sample user transaction from the sample system risk control decision graph based on the sample risk control system processing information, and determines a sample risk control path decision analysis based on the sample transaction risk control decision path, determines a first model loss value based on the "sample transaction risk control decision path and transaction risk control decision path label" using a model loss function, determines a second model loss value based on the "sample risk control path decision analysis and risk control path decision analysis label" using a model loss function, and obtains a model loss value of the initial system decision explanation model after the first model loss value and the second model loss value are combined, adjusts the model parameters of the initial system decision explanation model based on the model loss value until the model training end condition is met to obtain the system decision explanation model.

[0110] Optionally, in the subsequent model training process, after the model loss value is determined, the model recognition loss is only used to adjust the model parameters of the risk control decision explanation module in the initial system decision explanation model, and the model module parameters of the large language generation module are controlled to be unchanged.

[0111] C4: generating a target transaction risk control decision snapshot based on the target transaction risk control decision path and the risk control path decision analysis.

[0112] Illustratively, the risk control path decision analysis is added to the target transaction risk control decision path, so that the target transaction risk control decision snapshot can be generated.

[0113] S304: showing the target transaction risk control decision snapshot to a server through the risk control decision flow system.

[0114] In one or more embodiments of the present disclosure, it is difficult for general customer service personnel to accurately locate the risk control decision reason and reply to the customer using the server. In the present disclosure, the risk control visualization processing method is used to query the risk control system processing information of the risk control decision flow system when processing the target user transaction, and then the target user transaction corresponding target risk control system processing information can be extracted from the risk control system processing information. Then, according to the target risk control system processing information, the target transaction risk control decision snapshot for the target user transaction is extracted from the system risk control decision graph. Through the target transaction risk control decision snapshot, the risk control explainability of the specific target user transaction is greatly enhanced and the granularity is higher, focusing on the specific user transaction dimension. Through the visualization means, the system risk control decision process of the target user transaction can be restored, and the intelligent degree of the risk control decision is improved.

[0115] In the following, the risk control visualization processing device provided by the present disclosure will be described in detail in combination with FIG. 6. It should be noted that the risk control visualization processing device shown in FIG. 6 is used to execute the method of the embodiments shown in FIGS. 1-5 of the present disclosure. For the convenience of description, only the parts related to the present disclosure are shown, and the specific technical details not disclosed are referred to the embodiments shown in FIGS. 1-5 of the present disclosure.

[0116] Please refer to FIG. 6, which shows a structural schematic diagram of the risk control visualization processing device of the present disclosure. The risk control visualization processing device 1 can be realized by software, hardware or a combination of both to become all or part of a device. According to some embodiments, the risk control visualization processing device 1 includes an element determination module 11, a decision generation module 12 and a transaction processing module 13, which are specifically used for: the element determination module 11, configured to obtain risk control decision event information and determine at least one visualization flow element corresponding to the risk control decision event information; the decision generation module 12, configured to generate risk control decisions based on each visualization flow element to obtain a risk control decision flow system; and the transaction processing module 13, configured to process risk control transactions based on the risk control decision flow system and show a system risk control decision graph through the risk control decision flow system.

[0117] Optionally, the element determination module 11 is configured to: perform event element analysis processing on the risk control decision event information by using a risk control decision processing model to obtain at least one visualization flow element; and wherein the risk control decision processing model is trained by a plurality of sample risk control decision event information of known visualization flow element labels.

[0118] Optionally, the risk control decision processing model is obtained based on a basic large language generation model after being adapted to a risk control decision processing scenario, and the element determination module 11 is configured to: determine a decision event analysis description word for the risk control decision event information, input the risk control decision event information and the decision event analysis description word into a risk control decision processing model, perform event element analysis processing on the risk control decision event information through the risk control decision processing model to obtain at least one visual flow element, and output the at least one visual flow element.

[0119] Optionally, the decision generation module 12 is configured to: determine a sequential flow element, a flow node element block, and element configuration information corresponding to the flow node element block from each visual flow element, the flow node element block including a start node element block, a task element block, a gateway element block, an event element block, and an end node element block; configure the flow node element block based on the sequential flow element and the element configuration information to obtain a system risk control decision graph; and perform system code compilation processing on the system risk control decision graph to generate a risk control decision flow system.

[0120] Optionally, the transaction processing module 13 is configured to: obtain user risk control transaction data of at least one user object in a risk control decision scenario, call the risk control decision flow system to perform risk control transaction decision on the user risk control transaction data to obtain risk control system processing information; obtain a system risk control decision graph corresponding to the risk control decision flow system, determine a system flow node decision result based on the risk control system processing information, add the system flow node decision result to the system risk control decision graph, and display the system risk control decision graph to a server.

[0121] Optionally, the transaction processing module 13 is configured to: in response to a transaction execution query operation of a target user transaction by a server, extract a target transaction risk control decision snapshot for the target user transaction from the system risk control decision graph based on the risk control system processing information, and display the target transaction risk control decision snapshot to the server through the risk control decision flow system.

[0122] Optionally, the transaction processing module 13 is configured to: determine a target transaction risk control decision path for the target user transaction from the system risk control decision graph based on the risk control system processing information, determine a risk control path decision analysis for the target transaction risk control decision path, and generate a target transaction risk control decision snapshot based on the target transaction risk control decision path and the risk control path decision analysis.

[0123] Optionally, the transaction processing module 13 is configured to: generate a decision description prompt word for the target user transaction; input the system risk control decision graph, the risk control system processing information, and the decision description prompt word into a system decision interpretation model, determine a target transaction risk control decision path for the target user transaction from the system risk control decision graph based on the risk control system processing information through the system decision interpretation model, and determine a risk control path decision analysis based on the target transaction risk control decision path, and output the target transaction risk control decision path and the risk control path decision analysis.

[0124] It should be noted that the risk control visualization processing device provided in the above embodiments is only used as an example for the division of the above functional modules. In actual applications, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the above-described functions. In addition, the risk control visualization processing device and the risk control visualization processing method provided in the above embodiments belong to the same concept, and the implementation process is described in detail in the method embodiments, which will not be described here.

[0125] The serial numbers of the above disclosure are only for description, and do not represent the pros and cons of the embodiments.

[0126] In one or more embodiments of the present disclosure, the service platform determines at least one visual process element corresponding to the risk control decision event information by acquiring the risk control decision event information, then performs risk control decision generation processing based on each visual process element to obtain a risk control decision process system, and then performs risk control transaction processing based on the risk control decision process system and displays a system risk control decision graph through the risk control decision process system. The system risk control decision graph is displayed to the customer service end, the user end, the supervision party, and other non-system development side risk control decision process systems, the risk control explainability is greatly enhanced, the system risk control decision process can be restored through the visualization means, and the intelligent degree of the risk control decision is improved.

[0127] The present disclosure also provides a computer storage medium, which can store a plurality of instructions, the instructions being adapted to be loaded and executed by a processor to perform the risk control visualization processing method of the embodiments shown in FIGS. 1-5, and the specific execution process can be referred to the specific description of the embodiments shown in FIGS. 1-5, which will not be described here.

[0128] The present disclosure also provides a computer program product, which stores at least one instruction, the at least one instruction being loaded and executed by the processor to perform the risk control visualization processing method of the embodiments shown in FIGS. 1-5, and the specific execution process can be referred to the specific description of the embodiments shown in FIGS. 1-5, which will not be described here.

[0129] Please refer to FIG. 7, which is a structural block diagram of an electronic device provided by an embodiment of the present disclosure. The electronic device in the present disclosure can include one or more of the following components: a processor 110, a memory 120, an input device 130, an output device 140, and a bus 150. The processor 110, the memory 120, the input device 130, and the output device 140 can be connected through the bus 150.

[0130] The processor 110 can include one or more processing cores. The processor 110 connects various parts within the terminal through various interfaces and lines, executes various functions of the terminal and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 120, and calling data stored in the memory 120. Alternatively, the processor 110 can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 110 can integrate a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU is mainly used to process the operating system, user interface, and application programs; the GPU is used to render and draw display content; and the modem is used to process wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 110, but can be realized by a separate communication chip.

[0131] The memory 120 can include a random access memory (RAM) and can also include a read-only memory (ROM). Alternatively, the memory 120 includes a non-transitory computer-readable storage medium. The memory 120 can be used to store instructions, programs, codes, code sets or instruction sets.

[0132] The input device 130 is configured to receive input instructions or data, and the input device 130 includes but is not limited to a keyboard, a mouse, a camera, a microphone, or a touch device. The output device 140 is configured to output instructions or data, and the output device 140 includes but is not limited to a display device and a speaker. In the embodiments of the present disclosure, the input device 130 can be a temperature sensor configured to obtain the operating temperature of the terminal. The output device 140 can be a speaker configured to output an audio signal.

[0133] In addition, those skilled in the art can understand that the structure of the terminal shown in the above-mentioned drawings does not constitute a limitation on the terminal, and the terminal can include more or fewer components than those shown in the drawings, or combine certain components, or different component arrangements. For example, the terminal also includes radio frequency circuitry, an input unit, a sensor, audio circuitry, a wireless fidelity (WIFI) module, a power supply, a Bluetooth module, and the like, which are not described herein.

[0134] In the embodiments of the present disclosure, the execution subject of each step can be the terminal introduced above. Alternatively, the execution subject of each step is an operating system of the terminal. The operating system can be an Android system, an IOS system, or other operating systems, and the embodiments of the present disclosure do not limit the operating system.

[0135] In the electronic device of FIG. 7, the processor 110 can be configured to invoke a program stored in the memory 120 and perform to implement the risk control visualization processing method as described in various method embodiments of the present disclosure.

[0136] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when the program is executed, the processes of the above-mentioned embodiments can be included. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0137] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data for analysis, stored data, displayed data, etc.), and signals involved in the embodiments of the present disclosure are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data need to comply with relevant laws, regulations, and standards of countries and regions. For example, the risk control decision event information and the visualization process elements involved in the present disclosure are obtained under sufficient authorization.

[0138] The above disclosure is only the preferred embodiments of the present disclosure, and of course cannot limit the scope of the rights of the present disclosure, so equivalent changes made according to the claims of the present disclosure still fall within the scope of the present disclosure.

Claims

1. A risk control visualization processing method, the method comprising: Obtaining risk control decision event information, and determining at least one visual process element corresponding to the risk control decision event information; Perform risk control decision generation processing based on each of the visual process elements to obtain a risk control decision process system; Risk control transactions are processed based on the risk control decision process system and a system risk control decision diagram is displayed through the risk control decision process system.

2. The method according to claim 1, wherein determining at least one visual process element corresponding to the risk control decision event information comprises: Using a risk control decision processing model to perform event element analysis on the risk control decision event information to obtain at least one visual process element; The risk control decision processing model is trained by sample risk control decision event information with multiple known visual process element labels.

3. According to the method of claim 2, the risk control decision processing model is obtained by adapting the risk control decision processing scenario based on the basic large language generative model. The risk control decision processing model is used to perform event element analysis on the risk control decision event information to obtain at least one visual process element, including: Determine a decision event parsing descriptor for the risk control decision event information, input the risk control decision event information and the decision event parsing descriptor into a risk control decision processing model, perform event element parsing processing on the risk control decision event information through the risk control decision processing model to obtain at least one visual process element, and output the at least one visual process element.

4. The method according to claim 2, wherein the risk control decision generation process based on each of the visual process elements is performed to obtain a risk control decision process system, comprising: Determine, from each of the visual process elements, a sequence flow element, a process node element block, and element configuration information corresponding to the process node element block, wherein the process node element block includes a start node element block, a task element block, a gateway element block, an event element block, and an end node element block; Based on the sequence flow elements, the element configuration information is used to configure the process node element block to obtain a system risk control decision diagram; The system risk control decision diagram is compiled into system code to generate a risk control decision process system.

5. The method according to claim 1, wherein the processing of risk control transactions based on the risk control decision process system and displaying a system risk control decision diagram through the risk control decision process system comprises: Obtain user risk control transaction data of at least one user object in a risk control decision scenario, and call the risk control decision process system to perform risk control transaction decision on the user risk control transaction data to obtain risk control system processing information; Obtain the system risk control decision diagram corresponding to the risk control decision process system, determine the system process node decision result based on the risk control system processing information, add the system process node decision result to the system risk control decision diagram, and display the system risk control decision diagram to the server.

6. The method according to claim 5, after obtaining the risk control system processing information of the risk control decision process system during the risk control transaction decision process, further comprising: In response to the server executing a query operation on the target user transaction, a target transaction risk control decision snapshot for the target user transaction is extracted from the system risk control decision graph based on the risk control system processing information, and the target transaction risk control decision snapshot is displayed to the server through the risk control decision process system.

7. The method according to claim 6, wherein extracting a target transaction risk control decision snapshot for the target user transaction from the system risk control decision graph based on the risk control system processing information comprises: Determining a target transaction risk control decision path for the target user transaction from the system risk control decision graph based on the risk control system processing information, and determining a risk control path decision analysis for the target transaction risk control decision path; A target transaction risk control decision snapshot is generated based on the target transaction risk control decision path and the risk control path decision analysis.

8. The method according to claim 7, wherein determining a target transaction risk control decision path for the target user transaction from the system risk control decision graph based on the risk control system processing information, and determining a risk control path decision resolution for the target transaction risk control decision path, comprises: Generate decision description prompt words for the target user transaction; The system risk control decision diagram, the risk control system processing information and the decision description prompt word are input into a system decision interpretation model. The target transaction risk control decision path for the target user transaction is determined from the system risk control decision diagram based on the risk control system processing information by the system decision interpretation model, and a risk control path decision analysis is determined based on the target transaction risk control decision path. The target transaction risk control decision path and the risk control path decision analysis are output.

9. A risk control visualization processing device, comprising: An element determination module is used to obtain risk control decision event information and determine at least one visual process element corresponding to the risk control decision event information; A decision generation module, configured to generate risk control decisions based on each of the visual process elements to obtain a risk control decision process system; The transaction processing module is used to perform risk control transaction processing based on the risk control decision process system and display the system risk control decision diagram through the risk control decision process system.

10. A computer storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the method steps according to any one of claims 1 to 8.

11. A computer program product, wherein the computer program product stores at least one instruction, wherein the at least one instruction is loaded by a processor and executes the method steps according to any one of claims 1 to 8.

12. An electronic device comprising: A processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the method steps according to any one of claims 1 to 8.

Images