Risk early warning method, device, equipment, medium and program product

Abstract

The invention provides a risk early warning method which can be applied to the technical field of artificial intelligence. The method comprises the following steps: acquiring authorization of a target customer to customer information; after the authorization of the target customer to the customer information is obtained, the customer information is obtained, and the customer information comprises account information and behavior information; product information of m products is obtained, and m is an integer larger than or equal to 2; a knowledge graph is constructed according to the account information, the behavior information and the product information, nodes of the knowledge graph are constructed according to the account information and the product information, and edges of the knowledge graph are constructed according to the behavior information; according to the knowledge graph, obtaining a risk prediction result of the target customer by using a pre-constructed risk prediction model; and formulating an early warning strategy according to a risk prediction result. The invention further provides a risk early warning device and equipment, a storage medium and a program product.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, specifically to a risk warning method, apparatus, device, medium, and program product. Background Technology

[0002] Currently, risk control systems on financial product trading platforms typically employ two main data processing and risk warning mechanisms. The first mechanism is based on static questionnaire rule mapping. The system first presents the user with a set of standardized questions, each associated with a set of options. The system calculates the sum of predefined scores corresponding to the user's selected options, and then uses a pre-defined mapping table (e.g., mapping a specific score range to risk levels R1-R5) to output a static risk rating. The risk warning trigger logic for this mechanism is fixed: the system only matches the risk level of a product with the user's risk rating result when the user submits a transaction request to purchase a product with a specific risk level. If the risk level exceeds a preset threshold, a warning is triggered. The second mechanism attempts to dynamically assess risk by analyzing the user's historical trading records, such as by calculating statistical indicators like the volatility of the user's holdings. While this method introduces time-series data, its effective operation highly depends on long-term, continuous, and complete structured historical trading data. For new users, low-frequency traders, or users who only browse, query, and simulate configurations on the platform without actual trading, the system cannot obtain sufficient data samples to build an effective analytical model, resulting in data sparsity issues.

[0003] In summary, the existing technologies have the following technical problems: First, the event triggering mechanism of the risk warning system is rigid and lagging, and the data processing flow is asynchronous with the user's real-time decision-making behavior flow, resulting in the system's response time being significantly later than the risk budding stage; Second, the system's data analysis model has specific dependencies on the type and quality of input data and cannot integrate multimodal and fragmented real-time behavioral data streams, resulting in blind spots in the early identification of potential user risks, and the overall system's risk perception sensitivity and timeliness are insufficient. Summary of the Invention

[0004] In view of the above problems, this application provides risk warning methods, devices, equipment, media and program products that enable proactive risk assessment, strong early identification of potential risks and accurate risk prediction.

[0005] According to a first aspect of this application, a risk warning method is provided, comprising: obtaining authorization from a target customer for customer information; obtaining the customer information after obtaining authorization from the target customer, wherein the customer information includes account information and behavioral information; obtaining product information for m products, wherein m is an integer greater than or equal to 2; constructing a knowledge graph based on the account information, the behavioral information, and the product information, wherein nodes of the knowledge graph are constructed based on the account information and the product information, and edges of the knowledge graph are constructed based on the behavioral information; obtaining a risk prediction result for the target customer based on the knowledge graph using a pre-constructed risk prediction model; and formulating a warning strategy based on the risk prediction result.

[0006] According to embodiments of this application, the account information includes at least: name, account number, held products, and risk preference authentication; the behavioral information includes at least: frequency of product purchases and browsing, duration of each browsing session, quick swipe prompts for browsing products, number of times repeatedly clicking between two products, and risk warning content for screenshots of products; and the product information includes at least: product category, rate of return, Sharpe ratio, maximum drawdown, and risk level.

[0007] According to an embodiment of this application, the step of constructing a knowledge graph based on the account information, the behavioral information, and the product information includes: using the account entity of the account information as a node of the knowledge graph, and the name, the account number, the held products, and the risk preference certification as node attributes; using the product entity of the product information as a node of the knowledge graph, and the product category, the rate of return, the Sharpe ratio, the maximum drawdown, and the risk level as node attributes; determining the edges between the customer and the m products based on the behavioral information and pre-set edge rules, and using the edge rules satisfied by the edges as edge attributes of the edges.

[0008] According to an embodiment of this application, the step of obtaining a risk prediction result for the target customer based on the knowledge graph and using a pre-built risk prediction model includes: determining n contradictory node pairs in the knowledge graph based on the node attributes of the account entity node, the node attributes of the product entity node, and the edge attributes, where n is an integer greater than or equal to 1; determining the risk facts of each contradictory node pair based on the node attributes and edge attributes of each of the n contradictory node pairs; and obtaining the risk prediction result for the target customer based on the risk facts of the n contradictory node pairs using the pre-built risk prediction model.

[0009] According to an embodiment of this application, the step of obtaining a risk prediction result for the target customer based on the risk facts of the n contradictory node pairs using a pre-built risk prediction model includes: calculating the final risk prediction value of the target customer based on the risk facts of the n contradictory node pairs using the pre-built risk prediction model; subtracting the final risk prediction value from the built-in risk preference threshold of the target customer using the risk prediction model to obtain the risk deviation value of the target customer; and determining the risk prediction result of the target customer based on a set threshold range satisfied by the risk deviation value, wherein the risk prediction result includes one of risk preference escalation tendency, risk preference decline tendency, and risk preference remaining unchanged.

[0010] According to an embodiment of this application, the step of pre-constructing a risk prediction model includes: configuring a mapping table between conflict categories and weights in the risk prediction model; configuring a mapping table between risk facts and risk values ​​in the risk prediction model; mapping the account information of the target customer to a customer feature vector in the risk prediction model, and calculating the risk preference threshold of the target customer based on the customer feature vector; configuring a calculation method for calculating the final value of the risk prediction and the risk deviation value in the risk prediction model; and configuring a mapping table between a set threshold range and the risk prediction result in the risk prediction model.

[0011] A second aspect of this application provides a risk warning device, comprising: a first acquisition module, configured to acquire authorization from a target customer for customer information; a second acquisition module, configured to acquire the customer information after obtaining authorization from the target customer for the customer information, wherein the customer information includes account information and behavioral information; a third acquisition module, configured to acquire product information of m products, wherein m is an integer greater than or equal to 2; a construction module, configured to construct a knowledge graph based on the account information, the behavioral information, and the product information, wherein nodes of the knowledge graph are constructed based on the account information and the product information, and edges of the knowledge graph are constructed based on the behavioral information; a prediction module, configured to obtain a risk prediction result for the target customer based on the knowledge graph and using a pre-constructed risk prediction model; and a formulation module, configured to formulate a warning strategy based on the risk prediction result.

[0012] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0013] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0014] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0015] According to some embodiments of this application, customer information is obtained after obtaining authorization from the target customer, wherein the customer information includes account information and behavioral information; product information of m products is obtained; a knowledge graph can be constructed based on the account information, behavioral information, and product information, wherein the nodes of the knowledge graph are constructed based on the account information and product information, and the edges of the knowledge graph are constructed based on the behavioral information; based on the knowledge graph, a risk prediction result of the target customer can be obtained using a pre-built risk prediction model; and an early warning strategy can be formulated based on the risk prediction result. This application eliminates the need to wait for a specific transaction completion event as the sole trigger for early warnings. It initiates risk assessment in real-time by analyzing behavioral data streams through a risk prediction model during the early decision-making stages, such as user browsing, comparison, and research (represented by account and behavioral information). This significantly reduces the response delay from risk nascentness to early warning feedback, solving the technical challenge of severely delayed warning timing. Simultaneously, by effectively integrating non-transactional probing behaviors (represented by account and behavioral information) into the analytical framework, it reduces the strong reliance on structured historical transaction data, enhancing the early identification capability of potential user risks in data-sparse scenarios and resolving the risk identification blind spots caused by incomplete data perception dimensions and insufficient real-time performance. Furthermore, by introducing knowledge graphs to deeply integrate and structure account, behavioral, and product information, this application constructs a graph data model that reflects the dynamic relationship between users and products in real time. This enables the risk prediction model to continuously analyze and calculate based on continuous, multi-dimensional behavioral data streams, improving the accuracy of risk prediction. Attached Figure Description

[0016] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0017] Figure 1 The illustrations depict application scenarios of risk warning methods, apparatus, devices, media, and program products according to embodiments of this application.

[0018] Figure 2 A flowchart illustrating a risk warning method according to an embodiment of this application is shown schematically.

[0019] Figure 3A schematic diagram of a knowledge graph according to an embodiment of this application is shown.

[0020] Figure 4 This schematic diagram illustrates the structural block diagram of a risk warning device according to an embodiment of this application;

[0021] Figure 5 A block diagram schematically illustrates an electronic device suitable for implementing a risk warning method according to an embodiment of this application. Detailed Implementation

[0022] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0023] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0024] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0025] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0026] In the technical solution of this application, the customer information (including but not limited to user personal information, account information, behavioral information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the customer or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for customers to choose to authorize or refuse.

[0027] In scenarios where automated decision-making utilizes customer information, the methods, devices, and systems provided in this application all offer users corresponding operation entry points, allowing customers to choose to agree to or reject the automated decision results; if the customer chooses to reject, the process proceeds to the expert decision-making stage. Here, "automated decision-making" refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests, or economic and credit status through computer programs, and then making a decision. Here, "expert decision-making" refers to the activity of making decisions by personnel who specialize in a particular field, possess specialized experience, knowledge, and skills, and have reached a certain level of professional expertise.

[0028] The existing technologies have the following technical problems: First, the event triggering mechanism of the risk warning system is rigid and lagging, and the data processing flow is asynchronous with the user's real-time decision-making behavior flow, resulting in the system's response time being significantly later than the risk budding stage; Second, the system's data analysis model has specific dependencies on the type and quality of input data, and cannot integrate multimodal and fragmented real-time behavioral data flow, resulting in blind spots in the early identification of potential user risks, and the risk perception sensitivity and timeliness of the entire system are insufficient.

[0029] Embodiments of this application provide a risk warning method, apparatus, electronic device, computer-readable storage medium, and computer program product. The risk warning method includes: obtaining authorization from a target customer for customer information; obtaining the customer information after obtaining authorization, wherein the customer information includes account information and behavioral information; obtaining product information for m products, where m is an integer greater than or equal to 2; constructing a knowledge graph based on the account information, behavioral information, and product information, wherein nodes of the knowledge graph are constructed based on the account information and product information, and edges of the knowledge graph are constructed based on the behavioral information; obtaining a risk prediction result for the target customer based on the knowledge graph using a pre-built risk prediction model; and formulating a warning strategy based on the risk prediction result.

[0030] It should be noted that the risk warning method, device, electronic device, computer-readable storage medium and computer program product of this application can be used in the field of artificial intelligence technology, or in any field other than artificial intelligence technology, such as the financial field. The field of this application is not limited here.

[0031] Figure 1 The illustration shows application scenarios of risk warning methods, apparatus, devices, media, and program products according to embodiments of this application.

[0032] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0033] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0034] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0035] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0036] It should be noted that the risk warning method provided in this application embodiment can generally be executed by server 105. Correspondingly, the risk warning device provided in this application embodiment can generally be set in server 105. The risk warning method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the risk warning device provided in this application embodiment can also be set in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0037] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0038] The following will be based on Figure 1 The described scene, through Figure 2 and Figure 3 The risk warning method according to the embodiments of this application will be described in detail.

[0039] Figure 2 A flowchart illustrating a risk warning method according to an embodiment of this application is shown.

[0040] like Figure 2 As shown, the risk warning method in this embodiment includes operations S210 to S260.

[0041] In operation S210, obtain the target customer's authorization for customer information.

[0042] In operation S220, customer information is obtained after obtaining authorization from the target customer for customer information, which includes account information and behavioral information.

[0043] In embodiments of this application, the customer's consent or authorization can be obtained before acquiring the customer's information. For example, a request to acquire customer information can be sent to the customer before operation S220. If the customer agrees or authorizes the acquisition of customer information, operation S220 is performed.

[0044] In the embodiments of this application, a corresponding operation entry point can be provided to the customer, allowing the customer to choose to agree to or reject the automated decision-making result. That is, before constructing a knowledge graph, predicting risk prediction results, and formulating early warning strategies based on customer information, instructions for agreeing or rejecting to construct the knowledge graph, predict risk prediction results, and formulate early warning processes can be obtained from the customer through the corresponding operation entry point. If the customer agrees to construct the knowledge graph, predict risk prediction results, and formulate early warnings, then the knowledge graph, risk prediction results, and early warning processes are performed on the customer information, i.e., operations S240 to S260 are executed. If the customer refuses to construct the knowledge graph, predict risk prediction results, and formulate early warnings, then the expert decision-making process is initiated.

[0045] In operation S230, product information for m products is obtained, where m is an integer greater than or equal to 2.

[0046] As some feasible methods, account information can include at least: name, account number, held products, and risk preference authentication. Behavioral information can include at least: frequency of product purchases and browsing, duration of each browsing session, quick swiping prompts, number of times repeatedly clicking between two products, and risk warning content of screenshotted products. Product information can include at least: product category, return rate, Sharpe ratio, maximum drawdown, and risk level. Therefore, this application, through multi-dimensional and detailed data definitions (specific content of account information, behavioral information, and product information), enables the knowledge graph to construct richer node and edge relationships, providing a multi-dimensional decision-making basis for risk prediction models, thereby obtaining more accurate risk prediction results. This makes the early warning strategies formulated based on the risk prediction results more constructive and improves the early warning effect.

[0047] When operating S240, a knowledge graph is constructed based on account information, behavior information, and product information. Specifically, nodes of the knowledge graph are constructed based on account information and product information, and edges of the knowledge graph are constructed based on behavior information.

[0048] As one possible implementation method, the steps for constructing a knowledge graph based on account information, behavioral information, and product information include: a first construction operation, a second construction operation, and a third construction operation. The knowledge graph constructed through the first, second, and third construction operations can be... Figure 3 As shown.

[0049] The first construction operation involves using the account entity of the account information as a node in the knowledge graph, and the name, account number, held products, and risk preference authentication as node attributes.

[0050] The second construction operation involves using the product entities of the product information as nodes in the knowledge graph, and the product category, rate of return, Sharpe ratio, maximum drawdown, and risk level as node attributes.

[0051] The third construction operation: Based on the behavioral information and the pre-defined connection rules, determine the edges between the customer and m products, and use the connection rules that the edges satisfy as the edge attributes of the edges.

[0052] In some examples, pre-defined connection rules may include purchasing products, browsing products more frequently than a first set threshold, browsing products for a single time exceeding the average browsing time of all products on the platform, prompting users to quickly swipe through products, repeatedly clicking between two products more than a second set threshold, and risk warnings for taking screenshots of products, etc.

[0053] The first, second, and third construction operations facilitate the construction of a knowledge graph based on account, behavioral, and product information. Simultaneously, the first and second construction operations solidify customer risk preferences and product quantifiable risk indicators (such as the Sharpe ratio) into node attributes. This allows for precise identification of fundamental contradictions at the attribute level (e.g., conservative customer nodes versus high-risk product nodes), rather than merely focusing on superficial behavior. The third construction operation transforms complex behaviors into computable edge attributes. For example, by setting quantifiable rules (such as thresholds) for ambiguous behaviors like browsing frequency and repeated clicks, the invisible psychological activities of customers (such as decision-making hesitation and risk-ignoring anxiety) are mapped into explicit and traceable edge attributes in the knowledge graph. This allows risk prediction models to automatically identify abnormal trends inconsistent with risk preferences during early stages of customer behavior, such as high-frequency browsing and repeated comparisons, without waiting for the final purchase decision, enabling significantly earlier risk intervention. The knowledge graph built using the first, second, and third construction operations makes it possible to trace any risk warning back to which customer attribute, through which specific behaviors, and which product attribute it is associated with. This clear data structure not only improves the interpretability of the warning but also lays a solid foundation for developing targeted warning strategies.

[0054] When operating S250, risk prediction results for target customers are obtained by using a pre-built risk prediction model based on a knowledge graph.

[0055] As an implementable example, the steps of obtaining risk prediction results for target customers based on a knowledge graph and utilizing a pre-built risk prediction model include a first determination operation, a second determination operation, and a third determination operation.

[0056] The first determination operation is to determine n pairs of contradictory nodes in the knowledge graph based on the node attributes of the account entity node, the node attributes of the product entity node, and the edge attributes, where n is an integer greater than or equal to 1.

[0057] In some examples, for instance, if the node attribute of an account entity indicates a conservative risk preference, and this node has an edge connection to a product entity node with a high-risk attribute, and the edge attribute indicates that the frequency of browsing this product exceeds a first set threshold, then the node of the account entity and the node of the product entity can be identified as a contradictory pair. The above is merely an example of determining a contradictory node pair in a knowledge graph based on the node attributes of the account entity node, the node attributes of the product entity node, and the edge attributes. It should not be construed as a limitation of this application. The contradiction in a contradictory node pair can also lie in other aspects, which will not be elaborated upon here.

[0058] The second determination operation is to determine the risk facts for each of the n conflicting node pairs based on the node attributes and edge attributes of each conflicting node pair.

[0059] In some examples, identifying contradictory node pairs allows us to determine the account entity node, product entity node, and the edge connecting them. This yields the node attributes of the corresponding account entity node, the node attributes of the corresponding product entity node, and the edge attributes of the corresponding edge. Risk facts can be extracted from these node and edge attributes. Examples of risk facts include: an account with a conservative risk certification viewing a product with a high-risk node attribute; an account with a conservative risk certification browsing a high-risk product for a duration exceeding the platform's average browsing time; and an account with a conservative risk certification and a balance less than the minimum purchase amount for a product not actually trading a high-risk product, possibly due to funding restrictions. These risk facts are merely illustrative and should not be construed as limiting this application. Risk facts can be other facts as well; the specific content of risk facts is not limited here.

[0060] The third step involves determining the risk prediction results for the target customer based on the risk facts of n contradictory node pairs, using a pre-built risk prediction model.

[0061] As an implementable approach, the steps of obtaining risk prediction results for target customers based on the risk facts of n contradictory node pairs in a pre-built risk prediction model include a first calculation operation, a second calculation operation, and a fourth determination operation.

[0062] The first calculation operation: The pre-built risk prediction model calculates the final value of the risk prediction for the target customer based on the risk facts of n contradictory node pairs.

[0063] The second calculation operation: The risk prediction model subtracts the risk preference threshold of the target customer from the final risk prediction value to obtain the risk deviation value of the target customer.

[0064] The fourth step is to determine the risk prediction result for the target customer based on the set threshold range that the risk deviation value meets. The risk prediction result includes one of the following: risk preference escalation tendency, risk preference decline tendency, and risk preference remaining unchanged.

[0065] In some examples, risk prediction models can have a built-in mapping between conflict categories and weights. The weight of each conflict node pair can be obtained by matching the conflict category to each pair. Similarly, risk prediction models can have a built-in mapping between risk facts and risk values. The weight of each conflict node pair can be obtained by matching the conflict category to the k-value. i Each risk fact is matched with a corresponding risk value, k. i Let be an integer greater than or equal to 1, and let i be an integer greater than or equal to 1 and less than or equal to n.

[0066] For example, there are two contradictory node pairs. The first is an account entity whose node attribute indicates a conservative risk preference, and this node has an edge connection with a product entity node whose node attribute is high-risk. The edge attribute is that the frequency of browsing this product is greater than a first set threshold. The second is an account entity node with a balance of 'a' (a is a natural number greater than or equal to 0) and a product entity node with a minimum purchase amount of 'b' (b is a natural number greater than or equal to 0, and b is greater than a). The edge attribute of this edge connection is that the browsing frequency is greater than a first set threshold. The risk facts of the first contradictory node pair include: an account with a conservative risk preference viewing a product with a high risk preference; an account with a conservative risk preference viewing a product with a high risk preference for a duration exceeding the platform's average browsing duration; and an account with a conservative risk preference and a balance less than the minimum purchase amount for the product not actually trading a product with a high risk preference, possibly due to funding restrictions. The risk facts of the second contradictory node pair include: the account entity node's node attribute showing a balance 'a' less than the entity node's node attribute indicating a minimum purchase amount of 'b'; and the edge attribute of the edge between these two nodes indicating a browsing frequency greater than a first set threshold. Of course, the above contradictory node pairs and the risk facts for each contradictory node pair are merely illustrative examples and should not be construed as limitations on this application.

[0067] At the point of contradiction k i k in a risk fact i When the value equals 1, the risk value corresponding to that single risk fact can be used as the predicted risk value for the pair of contradictory nodes; in the case of k in the pair of contradictory nodes... i k in a risk fact iIf the value is greater than 1, then k can be... i The risk values ​​corresponding to each risk fact are summed to obtain the predicted risk value for that pair of contradictory nodes. A weighted average of the predicted risk values ​​for n pairs of contradictory nodes yields the final predicted risk value for the target customer. The risk prediction model incorporates a risk preference threshold for the target customer. The difference between the target customer's risk preference threshold and the final predicted risk value gives the target customer's risk deviation value. The predicted risk result for the target customer is determined based on the set threshold range satisfied by the risk deviation value. For example, if the risk deviation value is within the set threshold range 1, the predicted risk result for the target customer is a tendency to escalate risk preferences; if the risk deviation value is within the set threshold range 2, the predicted risk result for the target customer is a unchanged risk preference; and if the risk deviation value is within the set threshold range 3, the predicted risk result for the target customer is a tendency to decrease risk preferences.

[0068] The first calculation operation, the second calculation operation, and the fourth determination operation facilitate the implementation of the step in the third operation where the pre-built risk prediction model obtains the risk prediction result of the target customer based on the risk facts of n contradictory node pairs.

[0069] The first, second, and third determination operations facilitate the process of obtaining risk prediction results for target customers based on knowledge graphs and pre-built risk prediction models. Furthermore, this application identifies contradictory node pairs and the risk facts associated with them, then uses the risk prediction model to quantify these behaviors into final risk prediction values ​​and final risk deviation values. By comparing the risk deviation value with a set threshold range, it can identify the escalation tendency of a customer's risk appetite before the customer even expresses a purchase intention or makes a purchase. This allows banks to intervene promptly during the customer's decision-making stage (such as when they are hesitating or comparing products), providing cooling-off period reminders or educational content. This transforms risk warning from interception at the time of transaction to guidance before decision-making, fundamentally extending reaction time and preventing impulsive transactions and subsequent customer complaints.

[0070] As some feasible approaches, the steps of pre-constructing a risk prediction model may include a first configuration operation, a second configuration operation, a third configuration operation, a fourth configuration operation, and a fifth configuration operation.

[0071] The first configuration operation is to configure the mapping relationship table between conflict categories and weights in the risk prediction model.

[0072] The second configuration operation is to configure the mapping table between risk facts and risk values ​​in the risk prediction model.

[0073] The third configuration operation: The risk prediction model maps the target customer's account information into a customer feature vector, and calculates the target customer's risk preference threshold based on the customer feature vector.

[0074] Fourth configuration operation: Configure the calculation method for the terminal value of risk prediction and the risk deviation value in the risk prediction model.

[0075] Fifth configuration operation: Configure the mapping relationship table between the set threshold range and the risk prediction results in the risk prediction model.

[0076] In some examples, the method for calculating the final risk prediction value can be designed using a mapping table between conflict categories and weights, and a mapping table between risk facts and risk values. When the number of risk facts in a conflict node pair is equal to one, the risk value corresponding to that single risk fact can be used as the predicted risk value for that conflict node pair. When the number of risk facts in a conflict node pair is greater than one, the risk values ​​corresponding to these multiple risk facts are added together to obtain the predicted risk value for that conflict node pair. After obtaining the predicted risk values ​​for n conflict node pairs, a weighted average can be taken to obtain the final risk prediction value for the target customer.

[0077] In some examples, the method for calculating the risk deviation value can be designed to use a risk preference threshold, and the risk deviation value of the target customer is obtained by subtracting the target customer's risk preference threshold from the final value of the risk prediction.

[0078] By pre-setting multiple explicit mapping tables (such as mapping between conflict categories and weights, and between risk facts and risk values) through the first, second, third, fourth, and fifth configuration operations, the model's decision-making process becomes transparent. Clear rules govern how each action (such as a quick swipe prompt) is quantified and how different conflicts (such as mismatch between risk preference and product level) are weighted, enhancing the model's interpretability and credibility. Simultaneously, banks can optimize the model by adjusting parameters in the mapping tables based on business experience without needing to retrain complex algorithms, significantly improving the system's practicality and maintainability. Furthermore, the risk prediction model transforms from a black box into a highly configurable, interpretable decision engine that accurately adapts to individual differences, thereby ensuring the accuracy and reliability of the early warning results.

[0079] When operating S260, early warning strategies are formulated based on risk prediction results.

[0080] In some examples, for risk prediction results indicating an escalation of risk appetite, a pop-up window can be used to remind the customer to fill out the risk questionnaire again, or a pop-up window can be used to ask the customer to confirm whether to upgrade their risk certification level. In response to the customer's confirmation, the risk certification level can be upgraded directly for the customer; in response to the customer's rejection, the customer's risk certification level will not be changed.

[0081] In some examples, for risk prediction results indicating a downward trend in risk appetite, a pop-up window can be used to remind the customer to fill out the risk questionnaire again, or a pop-up window can be used to ask the customer to confirm whether to lower their risk certification level. In response to the customer's confirmation, the risk certification level can be lowered directly for the customer; in response to the customer's rejection, the customer's risk certification level will not be changed.

[0082] In some examples, for risk predictions where risk preference remains unchanged, it may not be necessary to alert the customer.

[0083] According to the risk warning method of this application embodiment, customer information is obtained after obtaining authorization from the target customer, wherein the customer information includes account information and behavioral information; product information of m products is obtained; a knowledge graph can be constructed based on the account information, behavioral information and product information, wherein the nodes of the knowledge graph are constructed based on the account information and product information, and the edges of the knowledge graph are constructed based on the behavioral information; based on the knowledge graph, the risk prediction result of the target customer can be obtained using a pre-constructed risk prediction model; and a warning strategy can be formulated based on the risk prediction result. This application eliminates the need to wait for a specific transaction completion event as the sole trigger for early warnings. It initiates risk assessment in real-time by analyzing behavioral data streams through a risk prediction model during the early decision-making stages, such as user browsing, comparison, and research (represented by account and behavioral information). This significantly reduces the response delay from risk nascentness to early warning feedback, solving the technical challenge of severely delayed warning timing. Simultaneously, by effectively integrating non-transactional probing behaviors (represented by account and behavioral information) into the analytical framework, it reduces the strong reliance on structured historical transaction data, enhancing the early identification capability of potential user risks in data-sparse scenarios and resolving the risk identification blind spots caused by incomplete data perception dimensions and insufficient real-time performance. Furthermore, by introducing knowledge graphs to deeply integrate and structure account, behavioral, and product information, this application constructs a graph data model that reflects the dynamic relationship between users and products in real time. This enables the risk prediction model to continuously analyze and calculate based on continuous, multi-dimensional behavioral data streams, improving the accuracy of risk prediction.

[0084] Based on the above-mentioned risk warning method, this application also provides a risk warning device. The following will combine... Figure 4 The device is described in detail.

[0085] Figure 4 A schematic block diagram of a risk warning device according to an embodiment of this application is shown.

[0086] like Figure 4 As shown, the risk warning device 10 includes a first acquisition module 1, a second acquisition module 2, a third acquisition module 3, a construction module 4, a prediction module 5, and a formulation module 6.

[0087] The first acquisition module 1 is used to obtain the target customer's authorization for customer information.

[0088] The second acquisition module 2 is used to acquire customer information after obtaining authorization from the target customer, wherein the customer information includes account information and behavioral information.

[0089] The third acquisition module 3 is used to acquire product information for m products, where m is an integer greater than or equal to 2.

[0090] Module 4 is used to construct a knowledge graph based on account information, behavior information, and product information. Specifically, it constructs the nodes of the knowledge graph based on account information and product information, and the edges of the knowledge graph based on behavior information.

[0091] Prediction module 5 is used to obtain risk prediction results for target customers based on the knowledge graph and a pre-built risk prediction model.

[0092] Module 6 is used to develop early warning strategies based on risk prediction results.

[0093] According to some embodiments of this application, the building module may include: a first building unit, a second building unit, and a third building unit.

[0094] The first building block uses account entities of account information as nodes in the knowledge graph, and name, account number, held products, and risk preference authentication as node attributes.

[0095] The second building block uses product entities from product information as nodes in the knowledge graph, and product category, rate of return, Sharpe ratio, maximum drawdown, and risk level as node attributes.

[0096] The third construction unit: Based on behavioral information and pre-defined connection rules, it determines the edges between the customer and m products, and uses the connection rules satisfied by the edge as the edge attribute of the edge.

[0097] According to some embodiments of this application, the prediction module may include a first determining unit, a second determining unit, and a third determining unit.

[0098] The first determining unit: Based on the node attributes of the account entity node, the node attributes of the product entity node, and the edge attributes, determine n pairs of contradictory nodes in the knowledge graph, where n is an integer greater than or equal to 1.

[0099] The second determining unit: Based on the node attributes and edge attributes of each of the n pairs of contradictory nodes, determine the risk facts of each pair of contradictory nodes.

[0100] The third determining unit: The pre-constructed risk prediction model obtains the risk prediction results for the target customer based on the risk facts of n contradictory node pairs.

[0101] According to some embodiments of this application, the third determining unit may include a first computing element, a second computing element, and a fourth determining element.

[0102] The first computational element: a pre-built risk prediction model calculates the final value of the target customer's risk prediction based on the risk facts of n contradictory node pairs.

[0103] The second calculation element: the risk prediction model calculates the difference between the built-in risk preference threshold of the target customer and the final value of the risk prediction to obtain the risk deviation value of the target customer.

[0104] The fourth determining element: Based on the set threshold range that the risk deviation value meets, determine the risk prediction result of the target customer, wherein the risk prediction result includes one of the following: risk preference escalation tendency, risk preference decline tendency, and risk preference unchanged.

[0105] According to some embodiments of this application, the risk warning device further includes a pre-construction module for pre-constructing a risk prediction model. The pre-construction module may include a first configuration unit, a second configuration unit, a third configuration unit, a fourth configuration unit, and a fifth configuration unit.

[0106] First configuration unit: Configure the mapping relationship table between contradiction categories and weights in the risk prediction model.

[0107] The second configuration unit: Configures the mapping table between risk facts and risk values ​​in the risk prediction model.

[0108] The third configuration unit: the risk prediction model maps the target customer's account information into a customer feature vector, and calculates the target customer's risk preference threshold based on the customer feature vector.

[0109] Fourth configuration unit: Configure the calculation methods for the final value of risk prediction and the risk deviation value in the risk prediction model.

[0110] Fifth configuration unit: Configure a mapping table between the set threshold range and the risk prediction results in the risk prediction model.

[0111] According to the risk warning device 10 of this application embodiment, customer information is obtained after obtaining authorization from the target customer, wherein the customer information includes account information and behavior information; product information of m products is obtained; a knowledge graph can be constructed based on the account information, behavior information and product information, wherein the nodes of the knowledge graph are constructed based on the account information and product information, and the edges of the knowledge graph are constructed based on the behavior information; based on the knowledge graph, the risk prediction result of the target customer can be obtained using a pre-constructed risk prediction model; and a warning strategy can be formulated based on the risk prediction result. This application eliminates the need to wait for a specific transaction completion event as the sole trigger for early warnings. It initiates risk assessment in real-time by analyzing behavioral data streams through a risk prediction model during the early decision-making stages, such as user browsing, comparison, and research (represented by account and behavioral information). This significantly reduces the response delay from risk nascentness to early warning feedback, solving the technical challenge of severely delayed warning timing. Simultaneously, by effectively integrating non-transactional probing behaviors (represented by account and behavioral information) into the analytical framework, it reduces the strong reliance on structured historical transaction data, enhancing the early identification capability of potential user risks in data-sparse scenarios and resolving the risk identification blind spots caused by incomplete data perception dimensions and insufficient real-time performance. Furthermore, by introducing knowledge graphs to deeply integrate and structure account, behavioral, and product information, this application constructs a graph data model that reflects the dynamic relationship between users and products in real time. This enables the risk prediction model to continuously analyze and calculate based on continuous, multi-dimensional behavioral data streams, improving the accuracy of risk prediction.

[0112] According to embodiments of this application, any multiple modules among the first acquisition module 1, second acquisition module 2, third acquisition module 3, construction module 4, prediction module 5, and formulation module 6 can be merged into one module, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this application, at least one of the first acquisition module 1, second acquisition module 2, third acquisition module 3, construction module 4, prediction module 5, and formulation module 6 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the first acquisition module 1, the second acquisition module 2, the third acquisition module 3, the construction module 4, the prediction module 5, and the formulation module 6 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0113] Figure 5 A block diagram schematically illustrates an electronic device suitable for the above-described method according to an embodiment of this application.

[0114] like Figure 5 As shown, an electronic device 900 according to an embodiment of this application includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage portion 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0115] RAM 903 stores various programs and data required for the operation of electronic device 900. Processor 901, ROM 902, and RAM 903 are interconnected via bus 904. Processor 901 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 902 and / or RAM 903. It should be noted that the programs may also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0116] According to embodiments of this application, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to a bus 904. The electronic device 900 may also include one or more of the following components connected to the input / output (I / O) interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the input / output (I / O) interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 910 as needed so that computer programs read from it can be installed into the storage section 908 as needed.

[0117] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0118] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 902 and / or RAM 903 and / or one or more memories other than ROM 902 and RAM 903 described above.

[0119] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the methods provided in the embodiments of this application.

[0120] When the computer program is executed by the processor 901, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0121] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 909, and / or installed from a removable medium 911. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0122] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 909, and / or installed from the removable medium 911. When the computer program is executed by the processor 901, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0123] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0124] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0125] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A risk warning method, characterized in that, include: Obtain authorization from target customers for customer information; The customer information is obtained after obtaining authorization from the target customer, and the customer information includes account information and behavioral information. Retrieve product information for m products, where m is an integer greater than or equal to 2; A knowledge graph is constructed based on the account information, the behavior information, and the product information, wherein the nodes of the knowledge graph are constructed based on the account information and the product information, and the edges of the knowledge graph are constructed based on the behavior information; Based on the knowledge graph, the risk prediction results for the target customer are obtained using a pre-built risk prediction model. Develop early warning strategies based on the risk prediction results.

2. The risk warning method according to claim 1, characterized in that, The account information includes at least: name, account number, products held, and risk preference verification; The behavioral information includes at least: the frequency of purchasing products, browsing products, the duration of each browsing session, prompts for quickly swiping while browsing products, the number of times products are repeatedly clicked between two products, and risk warnings for screenshots of products. The product information shall include at least the following: product category, yield, Sharpe ratio, maximum drawdown, and risk level.

3. The risk warning method according to claim 2, characterized in that, The steps of constructing a knowledge graph based on the account information, the behavioral information, and the product information include: The account entities of the account information are used as nodes in the knowledge graph, and the name, account number, held products, and risk preference authentication are used as node attributes. The product entities in the product information are used as nodes in the knowledge graph, and the product category, the rate of return, the Sharpe ratio, the maximum drawdown, and the risk level are used as node attributes. Based on the behavioral information and the pre-defined connection rules, the edges between the customer and the m products are determined, and the connection rules satisfied by the edges are taken as the edge attributes of the edges.

4. The risk warning method according to claim 3, characterized in that, The steps of obtaining risk prediction results for the target customer based on the knowledge graph and using a pre-built risk prediction model include: Based on the node attributes of the account entity node, the node attributes of the product entity node, and the edge attributes, n contradictory node pairs in the knowledge graph are determined, where n is an integer greater than or equal to 1. Based on the node attributes and edge attributes of each of the n pairs of contradictory nodes, determine the risk facts of each pair of contradictory nodes; The pre-built risk prediction model obtains the risk prediction result for the target customer based on the risk facts of the n contradictory node pairs.

5. The risk warning method according to claim 4, characterized in that, The steps of obtaining the risk prediction result for the target customer based on the risk facts of the n contradictory node pairs using a pre-built risk prediction model include: The pre-built risk prediction model calculates the final risk prediction value for the target customer based on the risk facts of the n contradictory node pairs; The risk prediction model calculates the risk deviation value of the target customer by subtracting the built-in risk preference threshold of the target customer from the final risk prediction value. The risk prediction result of the target customer is determined based on a set threshold range satisfied by the risk deviation value, wherein the risk prediction result includes one of the following: risk preference escalation tendency, risk preference decline tendency, and risk preference remaining unchanged.

6. The risk warning method according to claim 5, characterized in that, The steps for pre-building a risk prediction model include: Configure a mapping table between conflict categories and weights in the risk prediction model; Configure a mapping table between risk facts and risk values ​​in the risk prediction model; The risk prediction model maps the account information of the target customer into a customer feature vector, and calculates the risk preference threshold of the target customer based on the customer feature vector; Configure the calculation methods for the predicted risk final value and the risk deviation value in the risk prediction model; Configure a mapping table between the set threshold range and the risk prediction results in the risk prediction model.

7. A risk warning device, characterized in that, include: The first acquisition module is used to acquire the target customer's authorization for customer information; The second acquisition module is used to acquire the customer information after obtaining authorization from the target customer for the customer information, wherein the customer information includes account information and behavioral information; The third acquisition module is used to acquire product information of m products, where m is an integer greater than or equal to 2; A construction module is configured to construct a knowledge graph based on the account information, the behavior information, and the product information, wherein the nodes of the knowledge graph are constructed based on the account information and the product information, and the edges of the knowledge graph are constructed based on the behavior information. A prediction module is used to obtain the risk prediction result of the target customer based on the knowledge graph and a pre-built risk prediction model. The formulation module is used to formulate early warning strategies based on the risk prediction results.

8. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

Images