Consultation processing method and device, storage medium and electronic equipment

Abstract

The invention discloses a consultation processing method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining risk control consultation information which is inputted by a user to a risk control interpretation agent service and aims at a transaction risk control system, controlling a risk control interpretation big language model based on the risk control consultation information to determine target risk control attribution tool information, and obtaining risk control decision interpretation information for the transaction risk control system based on the target risk control attribution tool information, and outputting the risk control decision interpretation information to the user based on the risk control interpretation agent service.

Description

Technical Field

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

[0002] Risk control (referred to as risk control) on platform transaction services such as consumer finance transaction services, online shopping transaction services, and express logistics transaction services is mainly implemented through pre-developed transaction risk control systems. Functions such as data collection, three-party service integration, and algorithm execution are integrated into the transaction risk control system. After the development of the transaction risk control system is completed, the transaction risk control system is equivalent to a black box for non-system developers. Non-system developers have risk control consulting needs for the transaction risk control system in actual applications. Summary of the invention

[0003] This specification provides a consultation processing method, device, storage medium and electronic device, and the technical solution is as follows:

[0004] In a first aspect, this specification provides a consultation processing method, the method comprising:

[0005] Obtaining risk control consulting information for the transaction risk control system input by the user to the risk control explanation agent service, wherein the risk control consulting information includes consulting information for the system risk control transaction status;

[0006] Based on the risk control consultation information, the risk control interpretation large language model is controlled to determine target risk control attribution tool information, and based on the target risk control attribution tool information, risk control decision interpretation information for the transaction risk control system is obtained;

[0007] Based on the risk control interpretation agent service, risk control decision interpretation information is output to the user.

[0008] In a second aspect, this specification provides a risk control explanation large language model training method, the method comprising:

[0009] Use the basic big language model to create the initial risk control explanation big language model in the risk control explanation scenario;

[0010] Obtain sample risk control consulting information for the transaction risk control system in the risk control interpretation scenario;

[0011] The sample risk control consulting information is used to perform model training on the initial risk control explanation big language model. During the model training process, the sample risk control attribution tool information is determined through the initial risk control explanation big language model, and the predicted risk control decision explanation information for the transaction risk control system is obtained based on the sample risk control attribution tool information. The model parameters of the initial risk control explanation big language model are adjusted based on the predicted risk control decision explanation information until the initial risk control explanation big language model completes model training to obtain the risk control explanation big language model.

[0012] In a third aspect, this specification provides a consultation processing device, the device comprising:

[0013] An acquisition module, used to acquire risk control consulting information for a transaction risk control system input by a user to a risk control explanation agent service, wherein the risk control consulting information includes consulting information for a system risk control transaction status;

[0014] A determination module, configured to control the risk control interpretation large language model based on the risk control consulting information to determine target risk control attribution tool information, and to obtain risk control decision interpretation information for the transaction risk control system based on the target risk control attribution tool information;

[0015] An output module is used to output risk control decision explanation information to the user based on the risk control explanation agent service.

[0016] In a fourth aspect, this specification provides a risk control interpretation large language model training device, the device comprising

[0017] A model creation module is used to create an initial risk control explanation big language model in a risk control explanation scenario using a basic big language model;

[0018] An information acquisition module, used to acquire sample risk control consulting information for the transaction risk control system in the risk control interpretation scenario;

[0019] A model training module is used to use the sample risk control consulting information to perform model training on the initial risk control explanation big language model, determine the sample risk control attribution tool information through the initial risk control explanation big language model during the model training process, and obtain the predicted risk control decision explanation information for the transaction risk control system based on the sample risk control attribution tool information, adjust the model parameters of the initial risk control explanation big language model based on the predicted risk control decision explanation information, until the initial risk control explanation big language model completes model training, and obtains the risk control explanation big language model.

[0020] In a fifth aspect, the present specification provides a computer storage medium, wherein the computer storage medium stores at least one instruction, wherein the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present specification.

[0021] In a sixth aspect, the present specification provides a computer program product, wherein the computer program product stores at least one instruction, wherein the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present specification.

[0022] In a seventh aspect, the present specification provides an electronic device, which may include: 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 of one or more embodiments of the present specification.

[0023] The beneficial effects brought by the technical solutions provided by some embodiments of this specification include at least:

[0024] In one or more embodiments of the present specification, the electronic device obtains the risk control consultation information for the transaction risk control system input by the user to the risk control explanation agent service, controls the risk control explanation large language model based on the risk control consultation information to determine the target risk control attribution tool information, and obtains the risk control decision explanation information for the transaction risk control system based on the target risk control attribution tool information, and can output the risk control decision explanation information to the user based on the risk control explanation agent service. By adopting the method of combining the risk control explanation agent service with the risk control explanation large language model, the automation and accuracy of risk control management are significantly improved, so that users can obtain intuitive and personalized risk control explanations through simple natural language input. In addition, it not only reduces the dependence on professional risk control attribution tool knowledge and improves user experience, but also speeds up the response speed of risk control decisions, effectively supporting fast and accurate risk control evaluation and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in this specification or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 It is a scenario diagram of a consultation processing system provided in this specification;

[0027] Figure 2 It is a flowchart of a consultation processing method provided in this manual;

[0028] Figure 3 It is a schematic diagram of a consultation processing interface provided in this manual;

[0029] Figure 4 It is a schematic diagram of another consultation processing interface provided in this manual;

[0030] Figure 5 It is a flowchart of a consultation processing method provided in this manual;

[0031] Figure 6 It is a flowchart of information generation provided in this manual;

[0032] Figure 7 This is a flowchart of a risk control explanation large language model training method provided in this manual;

[0033] Figure 8 It is a schematic diagram of the device structure of a consultation processing device provided in this specification;

[0034] Fig. 9 It is a structural diagram of a risk control interpretation large language model training device provided in this specification;

[0035] Fig.10 It is a structural schematic diagram of an electronic device provided in this manual. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in this specification to clearly and completely describe the technical solutions in this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this specification.

[0037] In the description of this specification, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In the description of this specification, it should be noted that, unless otherwise clearly specified and limited, "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices. For those of ordinary skill in the art, the specific meanings of the above terms in this specification can be understood in specific circumstances. In addition, in the description of this specification, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are an "or" relationship.

[0038] In the relevant technology, non-system developers such as those on the client side, customer service side, user side, and regulatory side have risk control consulting needs for transaction risk control systems in actual applications. They need to attribute risk control anomalies to platform service transactions maintained by the transaction risk control system. Risk control anomaly attribution is a relatively complex proposition. In the daily work process, non-system developers such as those on the client side, customer service side, user side, and regulatory side need to consult system developers (such as risk control experts). System developers occupy a large amount of platform human resources and time costs because they need to frequently answer various risk anomaly questions.

[0039] The present specification is described in detail below with reference to specific embodiments.

[0040] See also Figure 1 , is a scenario diagram of a consultation processing system provided in this specification. Figure 1 As shown, the consultation processing system may include at least a client cluster and a service platform 100 .

[0041] The client cluster may include at least one client, such as Figure 1 As shown, it specifically includes client 1 corresponding to user 1, client 2 corresponding to user 2, ..., client n corresponding to user n, where n is an integer greater than 0.

[0042] Each client in the client cluster may be an electronic device with communication function, including but not limited to: wearable device, handheld device, personal computer, tablet computer, vehicle-mounted device, smart phone, computing device or other processing device connected to a wireless modem, etc. Electronic devices may 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.

[0043] The service platform 100 can be a separate server device, such as a rack-mounted, blade, tower, or cabinet-mounted server device, or a workstation, mainframe computer, or other hardware device with strong computing power; it can also be a server cluster composed of multiple servers, and the servers in the service cluster can be composed in a symmetrical manner, wherein each server has equivalent functions and status in the transaction link, and each server can provide services to the outside independently, and the independent service can be understood as not requiring the assistance of other servers.

[0044] In one or more embodiments of the present specification, the service platform 100 may establish a communication connection with at least one client in the client cluster, and complete the data interaction during the consultation process based on the communication connection.

[0045] 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, wherein 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, and the wired network includes but is not limited to Ethernet, a universal serial bus (USB) or a controller local 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), etc. are used to represent data (such as a target compressed package) exchanged through a network. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above data communication technologies.

[0046] The consultation processing system embodiment provided in this specification and the consultation processing method in one or more embodiments belong to the same concept. The execution subject corresponding to the consultation processing method involved in one or more embodiments of the specification can be an electronic device, and the electronic device can be the above-mentioned service platform 100; the execution subject corresponding to the consultation processing method involved in one or more embodiments of the specification can also be an electronic device corresponding to the client, which is determined based on the actual application environment. The embodiment of the consultation processing system and its implementation process can be detailed in the following method embodiment, which will not be repeated here.

[0047] based on Figure 1 The scenario diagram is shown, and the consultation processing method provided by one or more embodiments of this specification is introduced in detail below.

[0048] See also Figure 2, a flowchart of a consultation processing method is provided for one or more embodiments of the present specification, the method can be implemented by a computer program and can be run on a consultation processing device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application. The consultation processing device can be an electronic device.

[0049] Specifically, the consultation process includes:

[0050] S102: Acquire risk control consulting information for the transaction risk control system input by the user to the risk control explanation agent service, wherein the risk control consulting information includes consulting information for the system risk control transaction status;

[0051] Risk control explanation is also the risk control decision attribution process: the transaction risk control decision results are affected by many aspects of information, including traffic distribution, feature data, strategy models, engineering links, etc. Generally, when the risk decision transaction made by the transaction risk control system for a certain transaction presents a system risk control transaction status such as an abnormal risk control status, the user may need to quickly locate the root cause of the system risk control transaction status. The process of finding the root cause of the transaction decision is also called risk control decision attribution.

[0052] System risk control transaction status sets the transaction risk control system based on the actual transaction application scenario;

[0053] In this specification, the risk control interpretation agent service (Agent service) is a risk control interpretation computer program with perception, reasoning, and decision-making based on electronic devices. The risk control interpretation agent service is the service carrier object of the agent agent, relying on the risk control interpretation large language model in this specification to provide users with risk control consulting information for the transaction risk control system and autonomous interpretation processing functions for risk control decisions of the transaction risk control system;

[0054] Optionally, the risk control explanation agent service may be a service control with a risk control explanation function that carries the agent service. The user may trigger the service control corresponding to the risk control explanation agent service and input risk control consulting information for the transaction risk control system in the agent service interface corresponding to the service control.

[0055] Non-system developers such as clients, customer service, users, and regulators may be users involved in one or more embodiments of this specification. In actual applications, there is a need for risk control consulting on transaction risk control systems. Users may need to attribute risk control anomalies to the system risk control transaction status of a certain transaction item in the transaction risk control system. Risk control anomaly attribution: Transaction risk control decision results are affected by many aspects of information, including traffic distribution, feature data, strategy models, engineering links, etc. Generally, when the risk decision made by the transaction risk control system for a certain transaction item is in an abnormal risk control state, users may need to quickly locate the risk control anomaly problem and find the root cause. The process of finding the root cause of the decision anomaly is also called anomaly attribution.

[0056] At this time, the user may need to attribute the risk control anomaly to the platform service transaction maintained by the transaction risk control system. Based on this, the consultation processing method involved in this manual can be adopted. The user explains the risk control consultation information input by the intelligent service for the transaction risk control system to the risk control through the device he holds. At this time, the electronic device can obtain the risk control consultation information;

[0057] For example, taking consumer finance transactions as an example, if a consumer credit item applied for by a certain user is in an abnormal risk control state, the transaction risk control system rejects the consumer credit item. At this time, the risk control consultation information for the abnormal risk control state can be output to the risk control explanation agent service: "Help me find out why a certain user's consumer credit item was rejected."

[0058] For another example, taking consumer finance as an example, the credit rate of the platform's large-scale consumer credit items is lower than the credit rate of ordinary travel in a certain period of time. At this time, the risk control consultation information for the reduced credit rate of the large-scale consumer credit items can be output to the risk control explanation agent service: "Help me find out why the credit rate of the large-scale consumer credit items in a certain period of time has decreased";

[0059] Optionally, risk control consultation information may refer to a user's inquiry or question about the risk control status of a transaction in the transaction risk control system. Risk control consultation information includes but is not limited to the following aspects:

[0060] 1. Transaction identification information

[0061] Users may specify the identifier of a specific transaction in the risk control consultation information, such as the transaction number, user ID or other relevant unique identifier, so that the intelligent service can accurately locate the specific transaction that the user is concerned about through the risk control interpretation large language model.

[0062] 2. Query information about transaction status

[0063] Users may have specific questions about the current risk control status of a transaction, such as why a transaction is marked as high risk, or why an account is restricted from operation, etc. This type of consultation requires the transaction risk control system to provide an explanation of the current risk control status of the transaction, which may include the reason for the risk control mark, the measures taken and their consequences.

[0064] 3. Detailed information on risk control decisions

[0065] Users may be interested in the basis for the risk control decisions made by the system, such as what kind of data analysis or historical information is used to make the risk control assessment. At this time, users usually expect the system to explain in detail the various decision factors considered in the decision-making process.

[0066] 4. Motivation and background of consultation

[0067] The motivation of the user inquiring is also important information. For example, they may want to know how to optimize operations to avoid future risk control problems, or they may have objections to the risk control results and want to appeal. Understanding the user's motivation and background can help the agent better customize the response strategy through the risk control interpretation large language model.

[0068] S104: controlling the risk control interpretation large language model based on the risk control consulting information to determine target risk control attribution tool information, and acquiring risk control decision interpretation information for the transaction risk control system based on the target risk control attribution tool information;

[0069] Exemplarily, the risk control consulting information of the user is received through the risk control interpretation large language model, and the target risk control attribution tool that is most suitable for the current transaction risk control system is determined through semantic analysis and intent recognition of the risk control consulting information, so as to generate the target risk control attribution tool information corresponding to the target risk control attribution tool, so as to facilitate the subsequent use of the target risk control attribution tool based on the target risk control attribution tool information to obtain risk control decision explanation information for the transaction risk control system, specific tools such as case attribution tools, volatility attribution tools, etc.

[0070] In actual applications, several types of risk control attribution tools are pre-configured for the transaction risk control system. Risk control attribution tools play a vital role in the risk control system. Their main function is to deeply analyze and explain specific risk control events or decisions after the transaction risk control system is launched, helping users understand the root causes of risks and the impact and effects of related decisions. The application of risk control attribution tools not only improves the transparency and efficiency of risk control, but also helps management formulate more scientific and reasonable risk control strategies.

[0071] Risk control attribution tools may include one or more combinations of case attribution tools, volatility attribution tools, risk interpretation tools, user profile tools, etc. based on the different types of actual system risk control affairs status;

[0072] The target risk control attribution tool information includes at least the risk control attribution tool that needs to be called currently, and in some embodiments, also includes the parameter query information required by the tool determined from the risk control consulting information (that is, the target risk control transaction query parameter);

[0073] For example, the case attribution tool can determine the reason for risk control rejection of a single user in a specific scenario based on the user ID or traceID. Tool application scenario: It can be used for risk control analysis and management of individual users.

[0074] For example, the volatility attribution tool specifies a time range and analyzes the root causes of risk fluctuations within that period. The tool is used in scenarios where it helps to understand the changes in risks within a specific period of time and conduct time series analysis.

[0075] For example, the risk interpretation tool provides a textual interpretation of the user's risk status through the user ID. The tool is used in the following scenarios: It is suitable for the risk control department to conduct manual review and quickly obtain detailed descriptions of the user's risk status.

[0076] For example, the user profile tool can query the user's historical risk event records through the user ID. The tool is used to collect and analyze the user's long-term risk performance to facilitate the establishment of a risk management profile.

[0077] Optionally, the risk control attribution tool configured by the transaction risk control system is usually complex in design and usually requires risk control analysts or managers with professional knowledge. Ordinary users cannot directly use it due to objective reasons such as high technical complexity of the tool and strong professionalism in tool programming. In this manual, a large language model and an intelligent agent service are creatively combined as an intermediary to simplify the user's operation process. The user only needs to enter routine consulting information. The intelligent agent service can provide a simple and intuitive user interface. The user only needs to enter the queried risk control consulting information in the user interface corresponding to the risk control interpretation intelligent agent service. The intelligent agent service first receives the risk control consulting information entered by the user, and then performs semantic analysis and intent recognition on the risk control consulting information through the large language model to obtain the analysis results. According to the analysis results, the intelligent agent service determines the most suitable risk control attribution tool. For example, if a user inquires about the risk control reasons for a certain transaction, the case attribution tool can be called.

[0078] Subsequently, the risk control interpretation large language model or risk control interpretation intelligent agent service will automatically operate the target risk control attribution tool based on the target risk control attribution tool information, and use the target risk control attribution tool to process complex data analysis and model calls. Users can obtain risk control decision interpretation information for the transaction risk control system without directly contacting these complex operations. This step is fully automated to ensure the accuracy and efficiency of the analysis, making risk control management more friendly and accessible to non-professional users.

[0079] Exemplarily, the following methods may be used:

[0080] A2: Input the risk control consultation information into the risk control interpretation large language model;

[0081] A4: Control the risk control interpretation big language model to determine the consulting semantic intent of the risk control consulting information, and perform function call tool identification based on the consulting semantic intent in the big model function call tool set to obtain target risk control attribution tool information.

[0082] Exemplarily, the risk control interpretation large language model performs semantic intent recognition based on risk control consultation information to obtain consultation semantic intent, for example, the consultation information input by the user is processed through the integrated large language model. The model analyzes the text and identifies keywords and phrases, such as "high risk" and "transaction", to determine the core content and intent of the user's consultation. Then, based on the analysis results, the model understands the user's query intent and determines the user's consultation semantic intent. For example, the consultation semantic intent may be the need for specific transaction explanations, risk control policy interpretations, or detailed explanations of risk control tags. Then, it is necessary to perform function call tool recognition based on the consultation semantic intent from the large model function call tool set to obtain the target risk control attribution tool to generate target risk control attribution tool information.

[0083] Indicatively, the risk control explanation big language model is associated with a function call toolset, namely function_call. During the model training phase of the risk control explanation big language model, the corpus and parameters of several risk control attribution tools are registered in advance into the function_call of the big model. During the actual application phase, once the target risk control attribution tool information is determined, the tool is automatically called, and the target risk control attribution tool is used to input relevant data and parameters in the risk control attribution tool information, such as user ID, transaction ID, time range, etc., to obtain the data analysis results of the transaction risk control system, and generate risk control decision explanation information based on the data analysis results.

[0084] Risk control decision explanatory information: The risk control attribution tool usually queries professional data analysis results. Based on the data analysis results, the results are parsed to generate a detailed explanation of the risk control decision, and the risk control decision explanatory information is obtained, which may include, for example, the basis for the decision, relevant risk factors and recommended response measures.

[0085] S106: Outputting risk control decision explanation information to the user based on the risk control explanation agent service.

[0086] Exemplarily, the risk control explanation large language model can transmit the risk control decision explanation information to the risk control explanation agent service, and the risk control explanation agent service outputs the risk control decision explanation information to the user on the user interface. The risk control decision explanation information is displayed in a clear and easy-to-understand format in the user interface to ensure that the user can understand the reasons and background of the risk control decision.

[0087] For example, please refer to Figure 3 , Figure 3 It is a schematic diagram of a consultation processing interface. Because a user applied for a consumer credit item in a specific scenario, the transaction risk control system rejected the consumer credit item. At this time, the user explained the abnormal risk control status to the risk control agent service (such as Figure 3 The risk control attribution robot shown) inputs, for example, consulting information, such as Figure 3 In the example shown in the figure, “Help me check why the Huabei credit (a consumer credit item) of user ID 208830228686252981 was rejected”, by executing the consultation and explanation method of this specification, the risk control consultation information is used to control the risk control explanation large language model to determine that the target risk control attribution tool information is the case attribution tool information, and the case attribution tool is used based on the case attribution tool information to obtain the risk control decision explanation information for the transaction risk control system, and then the risk control explanation intelligent agent service outputs the following information to the user on the user interface: Figure 3 The risk control decision explanation information of the dialog box shown;

[0088] For example, please refer to Figure 4 , Figure 4 is a schematic diagram of another consultation processing interface. The credit rate of the platform's large-scale consumer credit items is lower than the credit rate of ordinary travel in a certain period of time. At this time, the user explains the lower credit rate of the large-scale consumer credit items to the risk control intelligent service (such as Figure 4 The risk control attribution robot shown) inputs, for example, consulting information, such as Figure 4 In the example shown in the figure, "Help me check why the approval rate of Huabei credit (a consumer credit item) has decreased in the past 10 minutes", by executing the consultation and interpretation method of this specification, the risk control consultation information is used to control the risk control interpretation large language model to determine that the target risk control attribution tool information is the volatility attribution tool information, and the volatility attribution tool is used based on the volatility attribution tool information to obtain the risk control decision interpretation information for the transaction risk control system, and then the risk control interpretation intelligent agent service outputs the following information to the user on the user interface: Figure 4 The risk control decision explanation information of the dialog box shown;

[0089] In one or more embodiments of the present specification, the electronic device obtains the risk control consultation information for the transaction risk control system input by the user to the risk control explanation agent service, controls the risk control explanation large language model based on the risk control consultation information to determine the target risk control attribution tool information, and obtains the risk control decision explanation information for the transaction risk control system based on the target risk control attribution tool information, and can output the risk control decision explanation information to the user based on the risk control explanation agent service. By adopting the method of combining the risk control explanation agent service with the risk control explanation large language model, the automation and accuracy of risk control management are significantly improved, so that users can obtain intuitive and personalized risk control explanations through simple natural language input. In addition, it not only reduces the dependence on professional risk control attribution tool knowledge and improves user experience, but also speeds up the response speed of risk control decisions, effectively supporting fast and accurate risk control evaluation and management.

[0090] See also Figure 5 , Figure 5 This is a flowchart of another embodiment of a consultation processing method proposed in one or more embodiments of this specification. Specifically:

[0091] S202: Inputting the risk control consultation information into the risk control interpretation language model;

[0092] S204: Controlling the risk control interpretation large language model to determine the consulting semantic intent of the risk control consulting information;

[0093] Exemplarily, after the risk control interpretation big language model determines the consulting semantic intent of the risk control consulting information, the target risk control attribution tool information is obtained by subsequently performing function call tool identification based on the consulting semantic intent in the big model function call tool set, and the target risk control attribution tool information includes the target risk control attribution tool and the target risk control transaction query parameters.

[0094] S206: Determine a target risk control attribution tool that matches the consulting semantic intent from the large model function call tool set, and obtain a target tool query parameter type corresponding to the target risk control attribution tool;

[0095] Exemplarily, the risk control interpretation large language model performs semantic intent recognition based on risk control consulting information to obtain consulting semantic intent, for example, the consulting information input by the user is processed through an integrated large language model. The model analyzes the text and identifies keywords and phrases, such as "high risk" and "transaction", to determine the core content and intent of the user's consultation. Then, based on the analysis results, the model understands the user's query intent and determines the user's consulting semantic intent. For example, the consulting semantic intent may be the need for specific transaction explanations, risk control policy interpretations, or detailed explanations of risk control tags. Then, it is necessary to perform function call tool recognition based on the consulting semantic intent from the large model function call tool set to obtain the target risk control attribution tool. In addition, it is necessary to obtain the tool query parameters required by the target risk control attribution tool, that is, to determine the target tool query parameter type corresponding to the target risk control attribution tool (pre-configured during the tool development phase);

[0096] S208: Extracting target risk control transaction query parameters corresponding to the target tool query parameter type based on the risk control consulting information.

[0097] For example, if the target tool query parameter type is user ID type, transaction ID type, time range type, etc., the user ID, transaction ID and time range are extracted based on the risk control consulting information to generate the target risk control transaction query parameter;

[0098] The risk control interpretation large language model generates target risk control attribution tool information based on the target risk control transaction query parameters and the target risk control attribution tool; subsequently, the target risk control attribution tool and the target risk control transaction query parameters can be determined based on the target risk control attribution tool information;

[0099] Then, based on the target risk control attribution tool information, risk control decision explanation information for the transaction risk control system is obtained.

[0100] S210: Based on the target risk control transaction query parameter, the risk control interpretation agent service is controlled to call the target risk control attribution tool to perform risk control decision information query processing to obtain target system risk control decision information;

[0101] Indicatively, the risk control explanation big language model is associated with function_call. During the model training phase of the risk control explanation big language model, the corpus and parameters of several risk control attribution tools are registered in advance into the function_call of the big model. Once the target risk control attribution tool information is determined during the actual application phase, the target risk control attribution tool is automatically called, and the target risk control transaction query parameters in the risk control attribution tool information, such as user ID, transaction ID, time range, etc., are input using the target risk control attribution tool to obtain the data analysis results of the transaction risk control system, and subsequently generate risk control decision explanation information based on the data analysis results.

[0102] S212: Generate risk control decision explanation information for the transaction risk control system based on the target system risk control decision information.

[0103] Exemplarily, the target system risk control decision information may be directly used as the risk control decision explanation information for the transaction risk control system;

[0104] Exemplarily, considering that the risk control attribution tool usually queries professional data analysis results, result analysis is performed based on the data analysis results to generate a detailed explanation of the risk control decision, and the risk control decision explanation information is obtained, which may include, for example, the basis for the decision, relevant risk factors and recommended response measures.

[0105] In one or more embodiments of the present specification, the electronic device obtains the risk control consultation information for the transaction risk control system input by the user to the risk control explanation agent service, controls the risk control explanation large language model based on the risk control consultation information to determine the target risk control attribution tool information, and obtains the risk control decision explanation information for the transaction risk control system based on the target risk control attribution tool information, and can output the risk control decision explanation information to the user based on the risk control explanation agent service. By adopting the method of combining the risk control explanation agent service with the risk control explanation large language model, the automation and accuracy of risk control management are significantly improved, so that users can obtain intuitive and personalized risk control explanations through simple natural language input. In addition, it not only reduces the dependence on professional risk control attribution tool knowledge and improves user experience, but also speeds up the response speed of risk control decisions, effectively supporting fast and accurate risk control evaluation and management.

[0106] Please refer to Figure 6 , Figure 6 It is a schematic diagram of the process of information generation. In a feasible implementation manner, the risk control decision explanation information for the transaction risk control system is generated based on the risk control decision information of the target system, and the following method can be adopted:

[0107] S3002: Determine a target risk control rule item based on the target system risk control decision information;

[0108] By analyzing the target system's risk control decision information, we can identify the target risk control rule items that specifically affect the decision, such as credit score thresholds, transaction behavior pattern matching, etc.

[0109] In schematic form, the key rules that currently affect specific risk control decisions in the transaction risk control system are identified and determined from the target system risk control decision information. These key rules, also known as the target risk control rule items, are the direct causes of the specific system risk control transaction status (such as transaction rejection, user risk rating adjustment, etc.).

[0110] S3004: performing a system rule operation evaluation on the transaction risk control system based on the target risk control rule item to obtain system rule operation evaluation information, and determining transaction optimization suggestion information for the target risk control rule item;

[0111] Exemplarily, after determining the key target risk control rule items, the effects of these target risk control rule items in actual operation are evaluated for system rule operation to determine whether the risk control decision system has false positives. The system rule operation evaluation includes analyzing the performance indicators such as the hit rate, false positive rate, and missed negative rate of the rules to evaluate their effectiveness and accuracy, thereby obtaining system rule operation evaluation information;

[0112] For example, based on the system rule operation evaluation information of rule operation, optimization measures for target risk control rule items are determined, and all optimization measures generate transaction optimization suggestion information. For example, if the false alarm rate of a rule is high, it may be recommended to adjust its parameters; or update the rule logic according to the latest risk data to improve the overall effectiveness and responsiveness of the system.

[0113] S3006: Generate risk control decision explanation information for the transaction risk control system based on the target system risk control decision information, the system rule operation evaluation information and the transaction optimization suggestion information.

[0114] Comprehensive information processing is performed based on the target system risk control decision information, the system rule operation evaluation information and the transaction optimization suggestion information: the risk control interpretation large language model combines the risk control decision information, the system rule operation evaluation results and the optimization suggestions to generate comprehensive risk control decision interpretation information, thereby providing users or decision makers with a clear and transparent risk control decision background view.

[0115] Output: Generated risk control decision explanation information, including detailed description of the role of each risk control rule, evaluation results and recommended optimization measures. In addition, it will explain how these rules work together to make specific risk control decisions, and how the optimization suggestions can improve system performance and decision quality.

[0116] In one or more embodiments of this specification, the risk control decision explanation information generated in the above manner can significantly improve the transparency and reliability of the risk control system, making the risk control decision process more reasonable and scientific. At the same time, through continuous rule evaluation and optimization, it can ensure that the risk control system keeps pace with the market environment and corporate strategy and effectively responds to various risk challenges. This method not only helps internal management, but also enhances user trust.

[0117] See also Figure 7, a flow chart of a risk control interpretation large language model training method is provided for one or more embodiments of this specification, the method can be implemented by a computer program and can be run on a consulting processing device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application. The consulting processing device can be an electronic device.

[0118] Specifically, the consultation process includes:

[0119] S4002: Using the basic big language model to create an initial risk control explanation big language model in the risk control explanation scenario;

[0120] In a feasible implementation, the risk control explanation large language model can be obtained by training the model for the risk control explanation scenario based on the basic large language model. The basic large language model (LLM) is an artificial intelligence content generation model that aims to understand and generate human language. The basic large language model is trained on a large amount of data and can perform a wide range of tasks, including text summarization, translation, sentiment analysis, etc.

[0121] Optionally, the basic large language model may be a Tongyi Qianwen large model, a Bailing large model, a GPT system large model, etc.;

[0122] In some embodiments, a trained basic large language model can be obtained, and the basic large language model can be adapted to the risk control explanation scenario to obtain the risk control explanation large language model. Usually, the basic large language model is directly applied to the dialogue response in the risk control explanation scenario, which is difficult to adapt to the new risk control explanation scenario. Based on this, the basic large language model is first obtained to create an initial risk control explanation large language model, and sample data in the new risk control explanation scenario is obtained. The sample data is sample risk control consulting information. Since the basic large language model is usually a trained AIGC model with content generation capabilities, based on this, this specification only needs to adapt it to the risk control explanation scenario. Specifically, the sample data can be used to perform model fine-tuning training on the initial risk control explanation large language model, and after the model fine-tuning training is completed, the risk control explanation large language model adapted to the risk control explanation scenario is obtained.

[0123] In a feasible implementation, since it is necessary to perform risk control interpretation on the risk control decision of the transaction risk control system, based on the model training phase of the large language model for risk control interpretation, the large model function call toolset is first configured, and the function call toolset is also function_call. In the model training phase of the large language model for risk control interpretation, the corpus and parameters of several risk control attribution tools are registered in advance in the function_call of the large model. Specifically, the following methods can be used:

[0124] B2: Obtain the risk control attribution query tool and tool query parameter type for the transaction risk control system in the risk control explanation scenario;

[0125] Risk control attribution tools may include one or more combinations of case attribution tools, volatility attribution tools, risk interpretation tools, user profile tools, etc. based on the different types of actual system risk control affairs status;

[0126] The tool query parameter type, that is, the parameter type that needs to be input when using the risk control attribution tool, is determined and configured in advance during the tool development phase.

[0127] B4: Use the basic big language model to create an initial risk control explanation big language model in the risk control explanation scenario, and configure the big model function calling toolset of the initial risk control explanation big language model based on the risk control attribution query tool and the tool query parameter type.

[0128] Exemplarily, model creation: obtaining a basic big language model, creating a risk control explanation scenario adaptation module under a risk control explanation scenario and using the basic big language model to create a basic big language model network, combining the risk control explanation scenario adaptation module and the basic big language model network to obtain an initial risk control explanation big language model; at the same time, based on the risk control attribution query tool and the tool query parameter type, a big model function call tool set for the initial risk control explanation big language model is configured in function_call;

[0129] Optionally, the risk control explanation scenario adaptation module can be created based on a machine learning model;

[0130] It should be noted that the machine learning models involved in one or more embodiments of the present specification include but are not limited to the fitting of one or more machine learning models such as convolutional neural network (CNN) model, deep neural network (DNN) model, recurrent neural network (RNN) model, embedding model, gradient boosting decision tree (GBDT) model, logistic regression (LR) model, etc.

[0131] S4004: Obtaining sample risk control consulting information for the transaction risk control system in the risk control interpretation scenario;

[0132] Sample data acquisition: obtain sample risk control consulting information under the new risk control interpretation scenario. The sample risk control consulting information can be the historical risk control consulting information of the platform;

[0133] Optional, sample data labeling: based on the risk control interpretation requirements of the risk control interpretation scenario, call the expert-side service to label the corresponding risk control decision interpretation information label.

[0134] S4006: Use the sample risk control consulting information to perform model training on the initial risk control explanation big language model, determine the sample risk control attribution tool information through the initial risk control explanation big language model during the model training process, and obtain the predicted risk control decision explanation information for the transaction risk control system based on the sample risk control attribution tool information, adjust the model parameters of the initial risk control explanation big language model based on the predicted risk control decision explanation information, until the initial risk control explanation big language model completes model training, and obtains the risk control explanation big language model.

[0135] In a feasible implementation manner, the determining of sample risk control attribution tool information by the initial risk control interpretation large language model during the model training process, and obtaining predicted risk control decision interpretation information for the transaction risk control system based on the sample risk control attribution tool information may be performed in the following manner:

[0136] C2: During the model training process, the initial risk control interpretation large language model is controlled to determine the sample consultation semantic intent of the sample risk control consultation information, and function call tool identification is performed based on the sample consultation semantic intent in the large model function call tool set to obtain predicted risk control attribution tool information;

[0137] C4: Determine the prediction risk control attribution tool and prediction risk control transaction query parameters based on the prediction risk control attribution tool information

[0138] C6: Based on the predicted risk control transaction query parameter, the risk control explanation agent service is controlled to call the predicted risk control attribution tool to perform risk control decision information query processing to obtain the predicted system risk control decision information;

[0139] C8: Generate forecast risk control decision explanation information for the transaction risk control system based on the forecast system risk control decision information.

[0140] The adjustment of model parameters of the initial risk control interpretation large language model based on the predicted risk control decision interpretation information in S4006 may be performed in the following manner:

[0141] The target model loss is determined based on the predicted risk control decision explanation information, the target model loss is used to adjust the model parameters of the risk control explanation scenario adaptation module, and the model parameters of the basic large language model network are controlled to remain unchanged.

[0142] Exemplarily, the model training process: input the sample risk control consulting information into the initial risk control explanation large language model for at least one round of model training to obtain predicted risk control decision explanation information, and use the model loss function to determine the target model loss based on the predicted risk control decision explanation information and the risk control decision explanation information label, and use the target model loss to adjust the model parameters of the risk control explanation scenario adaptation module, and control the model parameters of the basic large language model network to remain unchanged until the model training end conditions are met to obtain the risk control explanation scenario adaptation module and the basic large language model network, complete the model fusion of the basic large language model network and the risk control explanation scenario adaptation module, and obtain a trained risk control explanation large language model.

[0143] Optionally, the model training end condition of the model may include, for example, the value of the loss function is less than or equal to a preset loss function threshold, the number of iterations reaches a preset number threshold, etc. The specific model training end condition can be determined based on actual conditions and is not specifically limited here.

[0144] Model fusion of the basic large language model network and the risk control explanation scenario adaptation module: the weight of each model structure layer of the risk control explanation scenario adaptation module is fused with the basic large language model network, and the model structure layer parameters of the target model structure layer are determined in the basic large language model network. The model structure layer weight of the risk control explanation scenario adaptation module can only correspond to the model structure layer weight in part of all model structure layers in the basic large language model network. The model structure layer parameters of this target model structure layer are fused with the model structure layer weight of the risk control explanation scenario adaptation module. The model structure layer weight of the risk control explanation scenario adaptation module can only correspond to the model structure layer weight in part of all model structure layers in the basic large language model network. The model structure layer parameters are updated based on the model structure layer weight for this part of the target model structure layer, and the parameter update process of all model structure layer weights is completed by analogy, thereby obtaining a risk control explanation large language model.

[0145] In one or more embodiments of this specification, the risk control interpretation large language model is trained in the above manner. Based on the risk control interpretation large language model, the transparency and reliability of the transaction risk control system can be significantly improved, making the risk control decision-making process more reasonable and scientific. At the same time, through continuous rule evaluation and optimization, it can ensure that the risk control system keeps pace with the market environment and corporate strategy, and effectively responds to various risk challenges. This method not only helps internal management, but also enhances user trust.

[0146] The following will be combined Figure 8 , the consultation processing device provided in this specification is introduced in detail. It should be noted that, Figure 8 The consulting processing device shown is used to execute this instruction Figure 1 to Figure 7 For the convenience of explanation, only the parts related to this specification are shown. For specific technical details not disclosed, please refer to this specification. Figure 1 to Figure 7The embodiment shown.

[0147] See also Figure 8 , which shows a schematic diagram of the structure of the consultation processing device of this specification. The consultation processing device 1 can be implemented as all or part of the device through software, hardware or a combination of both. According to some embodiments, the consultation processing device 1 includes an acquisition module 11, a determination module 12 and an output module 13, which are specifically used to:

[0148] The acquisition module 11 is used to acquire the risk control consulting information for the transaction risk control system input by the user to the risk control explanation agent service, wherein the risk control consulting information includes consulting information for the system risk control transaction status;

[0149] A determination module 12 is used to control the risk control interpretation large language model based on the risk control consulting information to determine target risk control attribution tool information, and to obtain risk control decision interpretation information for the transaction risk control system based on the target risk control attribution tool information;

[0150] The output module 13 is used to output risk control decision explanation information to the user based on the risk control explanation agent service.

[0151] Optionally, the determining module 12 is used to:

[0152] Inputting the risk control consultation information into the risk control interpretation large language model;

[0153] The risk control interpretation big language model is controlled to determine the consulting semantic intent of the risk control consulting information, and function calling tool identification is performed based on the consulting semantic intent in the big model function calling tool set to obtain target risk control attribution tool information.

[0154] Optionally, the target risk control attribution tool information includes a target risk control attribution tool and a target risk control transaction query parameter, and the determination module 12 is used to:

[0155] Determine a target risk control attribution tool that matches the consulting semantic intent from the large model function call tool set, and obtain a target tool query parameter type corresponding to the target risk control attribution tool;

[0156] The target risk control transaction query parameter corresponding to the target tool query parameter type is extracted based on the risk control consulting information.

[0157] Optionally, the determining module 12 is used to:

[0158] Determine a target risk control attribution tool and a target risk control transaction query parameter based on the target risk control attribution tool information;

[0159] Based on the target risk control transaction query parameter, the risk control interpretation agent service is controlled to call the target risk control attribution tool to perform risk control decision information query processing to obtain the target system risk control decision information;

[0160] Based on the target system risk control decision information, risk control decision explanation information for the transaction risk control system is generated.

[0161] Optionally, the determining module 12 is used to:

[0162] Determine a target risk control rule item based on the target system risk control decision information;

[0163] Based on the target risk control rule item, the transaction risk control system is evaluated for system rule operation to obtain system rule operation evaluation information, and transaction optimization suggestion information for the target risk control rule item is determined;

[0164] Based on the target system risk control decision information, the system rule operation evaluation information and the transaction optimization suggestion information, risk control decision explanation information for the transaction risk control system is generated.

[0165] It should be noted that the consultation processing device provided in the above embodiment only uses the division of the above functional modules as an example when executing the consultation processing method. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the consultation processing device provided in the above embodiment belongs to the same concept as the consultation processing method embodiment. The implementation process thereof is detailed in the method embodiment and will not be repeated here.

[0166] The above serial numbers in this specification are for description only and do not represent the advantages or disadvantages of the embodiments.

[0167] In one or more embodiments of the present specification, the electronic device obtains the risk control consultation information for the transaction risk control system input by the user to the risk control explanation agent service, controls the risk control explanation large language model based on the risk control consultation information to determine the target risk control attribution tool information, and obtains the risk control decision explanation information for the transaction risk control system based on the target risk control attribution tool information, and can output the risk control decision explanation information to the user based on the risk control explanation agent service. By adopting the method of combining the risk control explanation agent service with the risk control explanation large language model, the automation and accuracy of risk control management are significantly improved, so that users can obtain intuitive and personalized risk control explanations through simple natural language input. In addition, it not only reduces the dependence on professional risk control attribution tool knowledge and improves user experience, but also speeds up the response speed of risk control decisions, effectively supporting fast and accurate risk control evaluation and management.

[0168] See also Fig. 9 , which shows a schematic diagram of the structure of the risk control interpretation large language model training device of this specification. The risk control interpretation large language model training device 2 can be implemented as all or part of the device through software, hardware or a combination of both. According to some embodiments, the risk control interpretation large language model training device 2 includes a model creation module 11, an information acquisition module 12 and a model training module 13, which are specifically used to:

[0169] A model creation module 11 is used to create an initial risk control explanation big language model in a risk control explanation scenario using a basic big language model;

[0170] An information acquisition module 12 is used to acquire sample risk control consulting information of the transaction risk control system in the risk control interpretation scenario;

[0171] The model training module 13 is used to use the sample risk control consulting information to perform model training on the initial risk control explanation big language model, determine the sample risk control attribution tool information through the initial risk control explanation big language model during the model training process, and obtain the predicted risk control decision explanation information for the transaction risk control system based on the sample risk control attribution tool information, adjust the model parameters of the initial risk control explanation big language model based on the predicted risk control decision explanation information, until the initial risk control explanation big language model completes model training, and obtains the risk control explanation big language model.

[0172] Optionally, the model creation module 11 is used to:

[0173] Obtain the risk control attribution query tool and tool query parameter type for the transaction risk control system in the risk control explanation scenario;

[0174] An initial risk control explanation big language model is created in a risk control explanation scenario using a basic big language model, and a big model function calling toolset of the initial risk control explanation big language model is configured based on the risk control attribution query tool and the tool query parameter type.

[0175] Optionally, the model training module 13 is used to:

[0176] During the model training process, the initial risk control interpretation large language model is controlled to determine the sample consultation semantic intent of the sample risk control consultation information, and function call tool identification is performed based on the sample consultation semantic intent in the large model function call tool set to obtain predicted risk control attribution tool information;

[0177] Determine the forecast risk control attribution tool and the forecast risk control transaction query parameter based on the forecast risk control attribution tool information;

[0178] Based on the predicted risk control transaction query parameter, the risk control interpretation agent service is controlled to call the predicted risk control attribution tool to perform risk control decision information query processing to obtain the predicted system risk control decision information;

[0179] Based on the forecasting system risk control decision information, forecasting risk control decision explanation information for the transaction risk control system is generated.

[0180] Optionally, the model training module 13 is used to:

[0181] Creating a risk control explanation scenario adaptation module in a risk control explanation scenario and using a basic large language model to create a basic large language model network, and obtaining an initial risk control explanation large language model based on the risk control explanation scenario adaptation module and the basic large language model network;

[0182] The adjusting model parameters of the initial risk control interpretation large language model based on the predicted risk control decision interpretation information includes:

[0183] The target model loss is determined based on the predicted risk control decision explanation information, the target model loss is used to adjust the model parameters of the risk control explanation scenario adaptation module, and the model parameters of the basic large language model network are controlled to remain unchanged.

[0184] It should be noted that the risk control interpretation large language model training device provided in the above embodiment only uses the division of the above-mentioned functional modules as an example when executing the risk control interpretation large language model training method. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the risk control interpretation large language model training device provided in the above embodiment and the risk control interpretation large language model training method embodiment belong to the same concept. The implementation process is detailed in the method embodiment, which will not be repeated here.

[0185] The above serial numbers in this specification are for description only and do not represent the advantages or disadvantages of the embodiments.

[0186] In one or more embodiments of the present specification, a trained risk control explanation large language model can be obtained through the above method, and the electronic device obtains the risk control consultation information for the transaction risk control system input by the user to the risk control explanation agent service, controls the risk control explanation large language model based on the risk control consultation information to determine the target risk control attribution tool information, and obtains the risk control decision explanation information for the transaction risk control system based on the target risk control attribution tool information, and then outputs the risk control decision explanation information to the user based on the risk control explanation agent service. By adopting the method of combining the risk control explanation agent service with the risk control explanation large language model, the automation and accuracy of risk control management are significantly improved, so that users can obtain intuitive and personalized risk control explanations through simple natural language input. In addition, it not only reduces the dependence on professional risk control attribution tool knowledge and improves user experience, but also speeds up the response speed of risk control decisions, effectively supporting fast and accurate risk control evaluation and management.

[0187] The present specification also provides a computer storage medium, which can store multiple instructions, which are suitable for being loaded and executed by a processor as described above. Figure 1 to Figure 7 The consultation processing method of the embodiment shown in the figure can be specifically executed by referring to Figure 1 to Figure 7 The specific description of the illustrated embodiment will not be repeated here.

[0188] The present specification also provides a computer program product, which stores at least one instruction, and the at least one instruction is loaded and executed by the processor as described above. Figure 1 to Figure 7 The consultation processing method of the embodiment shown in the figure can be specifically executed by referring to Figure 1 to Figure 7 The specific description of the illustrated embodiment will not be repeated here.

[0189] Please refer to Fig.10 , is a block diagram of a structure of an electronic device provided in an embodiment of this specification. The electronic device in this specification may include one or more of the following components: a processor 1010, a memory 1020, an input device 1030, an output device 1040, and a bus 1050. The processor 1010, the memory 1020, the input device 1030, and the output device 1040 may be connected via a bus 1050.

[0190] The processor 1010 may include one or more processing cores. The processor 1010 uses various interfaces and lines to connect various parts of the entire terminal, and executes various functions and processes data of the terminal 100 by running or executing instructions, programs, code sets or instruction sets stored in the memory 1020, and calling data stored in the memory 1020. Optionally, the processor 1010 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 1010 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 1010, but may be implemented separately through a communication chip.

[0191] The memory 1020 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 1020 includes a non-transitory computer-readable storage medium. The memory 1020 may be used to store instructions, programs, codes, code sets, or instruction sets.

[0192] The input device 1030 is used to receive input instructions or data, and the input device 1030 includes but is not limited to a keyboard, a mouse, a camera, a microphone, or a touch device. The output device 1040 is used to output instructions or data, and the output device 1040 includes but is not limited to a display device and a speaker. In the embodiment of this specification, the input device 1030 can be a temperature sensor for obtaining the operating temperature of the terminal. The output device 1040 can be a speaker for outputting audio signals.

[0193] In addition, those skilled in the art will appreciate that the structure of the terminal shown in the above drawings does not constitute a limitation on the terminal, and the terminal may include more or fewer components than shown, or combine certain components, or arrange the components differently. For example, the terminal also includes a radio frequency circuit, an input unit, a sensor, an audio circuit, a wireless fidelity (WIFI) module, a power supply, a Bluetooth module, and other components, which will not be described in detail here.

[0194] In the embodiments of this specification, the execution subject of each step may be the terminal described above. Optionally, the execution subject of each step is the operating system of the terminal. The operating system may be an Android system, an IOS system, or other operating systems, which is not limited in the embodiments of this specification.

[0195] exist Fig.10 In an electronic device, the processor 1010 can be used to call the program stored in the memory 1020 and execute it to implement the consultation processing method and / or the risk control interpretation large language model training method as described in the various method embodiments of this specification.

[0196] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0197] 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 used for analysis, stored data, displayed data, etc.) and signals involved in the embodiments of this specification are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions. For example, the risk control consulting information involved in this specification is obtained with full authorization.

[0198] The above disclosure is only the preferred embodiment of this specification, which certainly cannot be used to limit the scope of rights of this specification. Therefore, equivalent changes made according to the claims of this specification are still within the scope covered by this specification.

Claims

1. A consultation processing method, the method comprising: Obtaining risk control consulting information for the transaction risk control system input by the user to the risk control explanation agent service, wherein the risk control consulting information includes consulting information for the system risk control transaction status; Based on the risk control consultation information, the risk control interpretation large language model is controlled to determine target risk control attribution tool information, and based on the target risk control attribution tool information, risk control decision interpretation information for the transaction risk control system is obtained; Based on the risk control interpretation agent service, risk control decision interpretation information is output to the user.

2. According to the method of claim 1, the step of controlling the risk control interpretation large language model based on the risk control consulting information to determine the target risk control attribution tool information comprises: Inputting the risk control consultation information into the risk control interpretation large language model; The risk control interpretation big language model is controlled to determine the consulting semantic intent of the risk control consulting information, and function calling tool identification is performed based on the consulting semantic intent in the big model function calling tool set to obtain target risk control attribution tool information.

3. According to the method of claim 2, the target risk control attribution tool information includes the target risk control attribution tool and the target risk control transaction query parameter, The function calling tool identification based on the consulting semantic intent in the large model function calling tool set obtains the target risk control attribution tool information, including: Determine a target risk control attribution tool that matches the consulting semantic intent from the large model function call tool set, and obtain a target tool query parameter type corresponding to the target risk control attribution tool; The target risk control transaction query parameter corresponding to the target tool query parameter type is extracted based on the risk control consulting information.

4. According to the method of claim 1, the step of obtaining risk control decision explanation information for the transaction risk control system based on the target risk control attribution tool information comprises: Determine a target risk control attribution tool and a target risk control transaction query parameter based on the target risk control attribution tool information; Based on the target risk control transaction query parameter, the risk control interpretation agent service is controlled to call the target risk control attribution tool to perform risk control decision information query processing to obtain the target system risk control decision information; Based on the target system risk control decision information, risk control decision explanation information for the transaction risk control system is generated.

5. According to the method of claim 4, generating risk control decision explanation information for the transaction risk control system based on the target system risk control decision information comprises: Determine a target risk control rule item based on the target system risk control decision information; Based on the target risk control rule item, the transaction risk control system is evaluated for system rule operation to obtain system rule operation evaluation information, and transaction optimization suggestion information for the target risk control rule item is determined; Based on the target system risk control decision information, the system rule operation evaluation information and the transaction optimization suggestion information, risk control decision explanation information for the transaction risk control system is generated.

6. A risk control explanation large language model training method, the method comprising: Use the basic big language model to create the initial risk control explanation big language model in the risk control explanation scenario; Obtain sample risk control consulting information for the transaction risk control system in the risk control interpretation scenario; The sample risk control consulting information is used to perform model training on the initial risk control explanation big language model. During the model training process, the sample risk control attribution tool information is determined through the initial risk control explanation big language model, and the predicted risk control decision explanation information for the transaction risk control system is obtained based on the sample risk control attribution tool information. The model parameters of the initial risk control explanation big language model are adjusted based on the predicted risk control decision explanation information until the initial risk control explanation big language model completes model training to obtain the risk control explanation big language model.

7. According to the method of claim 6, the step of using the basic large language model to create an initial risk control interpretation large language model in a risk control interpretation scenario comprises: Obtain the risk control attribution query tool and tool query parameter type for the transaction risk control system in the risk control explanation scenario; An initial risk control explanation big language model is created in a risk control explanation scenario using a basic big language model, and a big model function calling toolset of the initial risk control explanation big language model is configured based on the risk control attribution query tool and the tool query parameter type.

8. The method according to claim 6, wherein determining sample risk control attribution tool information by using the initial risk control interpretation large language model during the model training process, and obtaining predicted risk control decision interpretation information for the transaction risk control system based on the sample risk control attribution tool information, comprises: During the model training process, the initial risk control interpretation large language model is controlled to determine the sample consultation semantic intent of the sample risk control consultation information, and function call tool identification is performed based on the sample consultation semantic intent in the large model function call tool set to obtain predicted risk control attribution tool information; Determine the forecast risk control attribution tool and the forecast risk control transaction query parameter based on the forecast risk control attribution tool information; Based on the predicted risk control transaction query parameter, the risk control interpretation agent service is controlled to call the predicted risk control attribution tool to perform risk control decision information query processing to obtain the predicted system risk control decision information; Based on the forecasting system risk control decision information, forecasting risk control decision explanation information for the transaction risk control system is generated.

9. According to the method of claim 6, the step of using the basic large language model to create an initial risk control interpretation large language model in a risk control interpretation scenario comprises: Creating a risk control explanation scenario adaptation module in a risk control explanation scenario and using a basic large language model to create a basic large language model network, and obtaining an initial risk control explanation large language model based on the risk control explanation scenario adaptation module and the basic large language model network; The adjusting model parameters of the initial risk control interpretation large language model based on the predicted risk control decision interpretation information includes: The target model loss is determined based on the predicted risk control decision explanation information, the target model loss is used to adjust the model parameters of the risk control explanation scenario adaptation module, and the model parameters of the basic large language model network are controlled to remain unchanged.

10. A consultation processing device, comprising: An acquisition module, used to acquire risk control consulting information for a transaction risk control system input by a user to a risk control explanation agent service, wherein the risk control consulting information includes consulting information for a system risk control transaction status; A determination module, configured to control the risk control interpretation large language model based on the risk control consulting information to determine target risk control attribution tool information, and to obtain risk control decision interpretation information for the transaction risk control system based on the target risk control attribution tool information; An output module is used to output risk control decision explanation information to the user based on the risk control explanation agent service.

11. A risk control interpretation large language model training device, the device comprising A model creation module is used to create an initial risk control explanation big language model in a risk control explanation scenario using a basic big language model; An information acquisition module, used to acquire sample risk control consulting information for the transaction risk control system in the risk control interpretation scenario; A model training module is used to use the sample risk control consulting information to perform model training on the initial risk control explanation big language model, determine the sample risk control attribution tool information through the initial risk control explanation big language model during the model training process, and obtain the predicted risk control decision explanation information for the transaction risk control system based on the sample risk control attribution tool information, adjust the model parameters of the initial risk control explanation big language model based on the predicted risk control decision explanation information, until the initial risk control explanation big language model completes model training, and obtains the risk control explanation big language model.

12. 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 10.

13. A computer program product, the computer program product storing 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 10.

14. 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 as claimed in any one of claims 1 to 10.

Images